Machine control using prediction maps

Abstract

One or more information maps are obtained by an agricultural work machine. The one or more information maps map one or more agricultural property values at different geographic locations of a field. An on-board sensor on the agricultural work machine senses the agricultural property as the agricultural work machine moves through the field. A prediction map generator generates a prediction map that predicts the agricultural property at different locations in the field based on a relationship between values in the one or more information maps and the agricultural property sensed by the on-board sensor. The prediction map can be output and used for automated machine control.

Description

[0001] This application is a continuation of, and claims priority to, Inventive Patent Application with application date February 7, 2021, application number 202110182559.0, entitled “Machine Control Using Prediction Maps”.

[0002] Cross Reference to Related Applications

[0003] This application is a continuation of, and claims priority to, U.S. Patent Application Serial No. 16 / 783,475 filed February 6, 2020, and U.S. Patent Application Serial No. 16 / 783,511 filed February 6, 2020, the contents of which are incorporated by reference herein in their entirety. TECHNICAL FIELD

[0004] This specification relates to agricultural machines, forestry machines, construction machines, and lawn care machines. BACKGROUND

[0005] There are various different types of agricultural machines. Some agricultural machines include harvesters, such as combine harvesters, sugar cane harvesters, cotton harvesters, self-propelled forage harvesters, and swathers. Some harvesters can be fitted with different types of headers to harvest different types of crops.

[0006] Weed clumps in a field have many detrimental effects on a harvesting operation. For example, when a harvester encounters a weed clump in a field, the weed clump can impede or cause the performance of the harvester to deteriorate. As a result, during a harvesting operation, when a weed clump is encountered, an operator can attempt to modify the controls of the harvester.

[0007] The above discussion is provided only as general background information and does not serve as an aid in determining the scope of the claimed subject matter. SUMMARY

[0008] One or more information maps are obtained by an agricultural work machine. The one or more information maps map one or more agricultural property values at different geographic locations of a field. An on-board sensor on the agricultural work machine senses the agricultural property as the agricultural work machine moves through the field. A prediction map generator generates a prediction map that predicts the agricultural property at different locations in the field based on a relationship between the values in the one or more information maps and the agricultural property sensed by the on-board sensor. The prediction map can be output and used for automatic machine control.

[0009] This Summary is provided to introduce some concepts in a simplified form that are further described below in the DETAILED DESCRIPTION. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to determine the scope of the claimed subject matter. The claimed subject matter is not limited to implementations that solve any or all disadvantages noted in the background. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 is a partial schematic diagram of an example of a combine harvester.

[0011] Figure 2 is a block diagram showing some parts of an agricultural harvester in more detail according to some examples of the present disclosure.

[0012] Figures 3A-3B is a flowchart showing an example of the operation of an agricultural harvester in generating a map.

[0013] Figure 4 is a block diagram showing one example of a prediction model generator and a prediction map generator.

[0014] Figure 5 is a flowchart showing an example of the operation of an agricultural harvester in receiving a vegetation index map, detecting weed characteristics, and generating a functional prediction weed map for controlling the agricultural harvester during a harvesting operation.

[0015] Figure 6A is a block diagram showing one example of a prediction model generator and a prediction map generator.

[0016] Figure 6B is a block diagram showing a field sensor.

[0017] Figure 7 is a flowchart showing one example of the operation of an agricultural harvester including generating a functional prediction map using a priori information map and field sensor input.

[0018] Figure 8 is a block diagram showing one example of a control zone generator.

[0019] Figure 9 is a flowchart showing one example of the operation of the control zone generator shown in Figure 8

[0020] Figure 10 is a flowchart showing one example of the operation of a control system in selecting a target setpoint value to control an agricultural harvester.

[0021] Figure 11 is a block diagram showing one example of an operator interface controller.

[0022] Figure 12 is a flowchart showing one example of an operator interface controller.

[0023] Figure 13 is a schematic diagram showing one example of an operator interface display.

[0024] Figure 14 ​is a block diagram illustrating one example of an agricultural harvester in communication with a remote server environment.

[0025] Figures 15-17 An example of a mobile device that can be used with an agricultural harvester is shown.

[0026] Figure 18 is a block diagram illustrating one example of a computing environment that can be used with an agricultural harvester. DETAILED DESCRIPTION

[0027] To facilitate an understanding of the principles of the present disclosure, reference is made to the examples illustrated in the drawings, and specific language will be used to describe the examples. It will, however, be understood that no limitation of the scope of the disclosure is intended. Alterations and further modifications of the described devices, systems, methods, and any further applications of the principles of the present disclosure are fully contemplated as would occur to one skilled in the art to which the present disclosure relates. In particular, it is fully contemplated that the features, components, and / or steps described with respect to one example can be combined with the features, components, and / or steps described with respect to other examples of the present disclosure.

[0028] This specification relates to using in-field data acquired contemporaneously with an agricultural operation in conjunction with prior data (previous data) to generate a prediction map, such as a prediction weed map. In some examples, the prediction map can be used to control an agricultural work machine, such as an agricultural harvester. As discussed above, performance of an agricultural harvester can be degraded when the agricultural harvester engages a weed mass. For example, if the crop in a field has ripened, the weeds present in the field can still be green, thus increasing the moisture content of the biomass encountered by the agricultural harvester. This problem can be exacerbated when the weed mass in the field is wet (such as shortly after a rain or when the weed mass contains dew) and before the weed mass has had a chance to dry.

[0029] The performance of an agricultural harvester can be adversely affected based on a number of different criteria. For example, the strength of weeds in a patch of weeds can have a deleterious effect on the operation of an agricultural harvester. Without limitation, weed strength can include at least one of the presence of weeds, weed population, weed growth stage, weed biomass, weed moisture, weed density, height of weeds, size of weed plants, age of weeds, or health of weeds at a location within an area. A measure of weed strength can be a binary value (such as weeds present or weeds not present), or a continuous value (such as a percentage of weeds in a defined area or volume), or a set of discrete values (such as a low, medium, or high weed strength value). Similarly, different types of weeds encountered by an agricultural harvester can affect the agricultural harvester differently. For example, different types of weeds can contain different levels of moisture, and as the level of moisture of weeds increases, the degradation in harvester performance can also increase. Similarly, different weed types can have different physical structures (e.g., some weeds can have thicker or thinner stems, wider leaves, etc.). These variations in weed structure can also result in variations in performance of an agricultural harvester as the agricultural harvester engages with such weeds.

[0030] Vegetation index maps illustratively map vegetation index values (which can be indicative of vegetation growth) at different geographic locations in a field of interest. One example of a vegetation index includes the normalized difference vegetation index (NDVI). Many other vegetation indices are also within the scope of the present disclosure. In some examples, a vegetation index can be derived from sensor readings of one or more bands of electromagnetic radiation reflected by a plant. Without limitation, these bands can be in the microwave, infrared, visible, or ultraviolet portions of the electromagnetic spectrum.

[0031] Vegetation index maps can be used to identify the presence and location of vegetation. In some examples, these maps enable identification and georeferencing of weeds in the presence of bare soil, crop residue, or other plant life, including crops or other weeds. For example, at the end of a growing season, when crops are mature, crop plants can exhibit a reduced level of live growing vegetation. However, weeds typically continue in a growing state after crops are mature. Thus, if a vegetation index map is generated relatively late in the growing season, the vegetation index map can indicate the location of weeds in a field. However, in certain cases, vegetation index maps can be less useful (or not useful at all) in identifying the strength of weeds in a patch of weeds or the type of weeds in a patch of weeds. Thus, in certain cases, vegetation index maps can have reduced usefulness in predicting how to control an agricultural harvester as it moves through a field.

[0032] Accordingly, the present discussion is directed to systems that receive an a priori information map of a field or a map generated during an a priori operation (prior operation) and also use in-field sensors to detect variables indicative of one or more of agricultural properties (such as biomass) during a harvesting operation, machine speed, or operator command inputs. The command inputs can be setting inputs or other control inputs set for controlling an agricultural harvester, such as steering inputs, speed inputs, header height inputs, and other inputs. The system generates a model that models the relationship between values on the a priori information map and output values from the in-field sensors. The model is used to generate a functional prediction map that predicts, for example, biomass, machine speed, or operator command inputs at different locations in the field. The functional prediction map generated during the harvesting operation can be presented to an operator or other user, or used to automatically control the agricultural harvester during the harvesting operation, or both. The functional prediction map can be used to control one or more of a feed rate, machine speed, and command inputs.

[0033] Figure 1 is a partial schematic view of a self-propelled agricultural harvester 100. In the example shown, the agricultural harvester 100 is a combine harvester. In addition, although a combine harvester is provided as an example throughout this disclosure, it should be understood that the present description also applies to other types of harvesters, such as cotton harvesters, sugar cane harvesters, self-propelled forage harvesters, swathers, or other agricultural work machines. Accordingly, the present disclosure is intended to encompass the various types of harvesters described and is therefore not limited to combine harvesters. Also, the present disclosure relates to other types of work machines, such as agricultural planters and sprayers, construction equipment, forestry equipment, and lawn care equipment, where the generation of a prediction map can be applied. Accordingly, the present disclosure is intended to encompass these various types of harvesters and other work machines and is therefore not limited to combine harvesters.

[0034] As Figure 1As shown, the agricultural harvester 100 illustratively includes an operator compartment 101 that can have various different operator interface mechanisms for controlling the agricultural harvester 100. The agricultural harvester 100 includes front end equipment such as a header 102 and a cutter 104 generally indicated at 104. In the illustrated example, the cutter 104 is included on the header 102. The agricultural harvester 100 also includes a feedhouse 106, a feed accelerator 108, and a threshing machine generally indicated at 110. The feedhouse 106 and the feed accelerator 108 form part of a material handling subsystem 125. The header 102 is pivotally coupled to a frame 103 of the agricultural harvester 100 along a pivot axis 105. One or more actuators 107 drive the header 102 to move about the axis 105 in a direction generally indicated by arrow 109. Thus, a vertical position (header height) of the header 102 above a ground surface 111 on which the header 102 travels is controllable by actuating the actuators 107. Although not shown in Figure 1 the agricultural harvester 100 can also include one or more actuators that operate to apply a tilt angle, a roll angle, or both, to the header 102 or portions of the header 102. Tilt refers to an angle at which the cutter 104 engages the crop. For example, the tilt angle is increased by controlling the header 102 to point a distal edge 113 of the cutter 104 more toward the ground surface. The tilt angle is decreased by controlling the header 102 to point the distal edge 113 of the cutter 104 more away from the ground surface. Roll angle refers to an orientation of the header 102 about a fore-aft longitudinal axis of the agricultural harvester 100.

[0035] The threshing machine 110 illustratively includes a threshing rotor 112 and a set of concaves 114. In addition, the agricultural harvester 100 also includes a separator 116. The agricultural harvester 100 also includes a grain cleaning subsystem or grain cleaning house (collectively referred to as a grain cleaning subsystem 118) that includes a grain cleaning fan 120, a chaffer 122, and a sieve 124. The material handling subsystem 125 also includes a beater threshing cylinder 126, a tailings elevator 128, a clean grain elevator 130, and an unloading auger 134 and a spout 136. The clean grain elevator moves clean grain into a clean grain tank 132. The agricultural harvester 100 also includes a residue subsystem 138 that can include a chopper 140 and a spreader 142. The agricultural harvester 100 also includes a propulsion subsystem that includes an engine that drives ground engaging components 144 (e.g., wheels or tracks). In some examples, a combine within the scope of the present disclosure can have more than one of any of the above-mentioned subsystems. In some examples, the agricultural harvester 100 can have a left grain cleaning subsystem and a right grain cleaning subsystem, a separator, etc., which are not shown in Figure 1

[0036] ​In operation, and as outlined, the agricultural harvester 100 illustratively moves through a field in a direction indicated by arrow 147. As the agricultural harvester 100 moves, the header 102 (and associated reel 164) engages the crop to be harvested and gathers the crop toward the cutter 104. The operator of the agricultural harvester 100 can be a local human operator, a remote human operator, or an automated system. Operator commands are commands issued by the operator. The operator of the agricultural harvester 100 can determine one or more of a height setting, a tilt angle setting, or a roll angle setting of the header 102. For example, the operator inputs one or more settings to a control system that controls the actuator 107 (described in more detail below). The control system can also receive settings from the operator to establish the tilt angle and roll angle of the header 102 and implement the input settings by controlling associated actuators (not shown) that operate to change the tilt angle and roll angle of the header 102. The actuator 107 maintains the header 102 at a height above the ground 111 based on the height setting, and, where applicable, at a desired tilt and roll angle. Each of the height, roll, and tilt settings can be implemented independently of the other settings. The control system responds to header errors (e.g., a difference between the height setting and a measured height of the header 104 above the ground 111, and in some examples, tilt angle and roll angle errors) with a responsiveness determined based on a selected level of sensitivity. If the level of sensitivity is set at a higher level of sensitivity, the control system responds to smaller header position errors and attempts to reduce detected errors more quickly than if the level of sensitivity is at a lower level of sensitivity.

[0037] Returning to the description of the operation of the agricultural harvester 100, after the crop is cut by the cutter 104, the cut crop material moves through the conveyor in the feedhouse 106 toward the feed accelerator 108, which accelerates the crop material into the threshing machine 110. The crop material is threshed by the rotor 112, which rotates the crop against the concaves 114. The threshed crop material is moved by a separator rotor in the separator 116, where a portion of the straw is moved by the unloading threshing cylinder 126 toward the straw sub-system 138. The portion of the straw that is conveyed to the straw sub-system 138 is chopped by the straw chopper 140 and spread on the field by the spreader 142. In other configurations, the straw is released from the agricultural harvester 100 into a pile. In other examples, the straw sub-system 138 can include a weed seed rejector (not shown), such as a seed bagger or other seed collector, or a seed crusher or other seed destroyer.

[0038] The grain falls to the grain cleaning subsystem 118. The chaffer 122 separates some of the larger pieces of material from the grain, and the sieve 124 separates some of the finer pieces of material from the clean grain. The clean grain falls onto an auger that moves the grain to the inlet end of the clean grain elevator 130, and the clean grain elevator 130 moves the clean grain upward, storing the clean grain in the clean grain bin 132. The residue is removed from the grain cleaning subsystem 118 by the airflow generated by the grain cleaning fan 120. The grain cleaning fan 120 directs air up through the sieve and the chaffer along an airflow path. The airflow carries the residue in the agricultural harvester 100 rearward toward the residue handling subsystem 138.

[0039] The tailings elevator 128 returns the tailings to the threshing machine 110, where the tailings are re-threshed. Alternatively, the tailings can also be passed to a separate re-threshing mechanism by the tailings elevator or another transport device, where the tailings are also re-threshed.

[0040] Figure 1 It is also shown that in one example, the agricultural harvester 100 includes a machine speed sensor 146, one or more separator loss sensors 148, a clean grain camera 150, a forward view image capture mechanism 151 that can be in the form of a stereo camera or a monocular camera, and one or more loss sensors 152 disposed in the grain cleaning subsystem 118.

[0041] The machine speed sensor 146 senses the speed of travel of the agricultural harvester 100 over the ground. The machine speed sensor 146 can sense the speed of travel of the agricultural harvester 100 by sensing the rotational speed of ground-engaging components such as wheels or tracks, drive shafts, axles, or other components. In some cases, a positioning system can be used to sense the speed of travel, such as a global positioning system (GPS), a dead reckoning system, a long range navigation (LORAN) system, or various other systems or sensors that provide an indication of the speed of travel.

[0042] The loss sensors 152 schematically provide output signals indicative of the amount of grain loss occurring in the right and left sides of the grain cleaning subsystem 118. In some examples, the sensors 152 are impact sensors that count the grain impacts per unit of time or per unit of travel distance to provide an indication of the grain loss occurring at the grain cleaning subsystem 118. The impact sensors for the right and left sides of the grain cleaning subsystem 118 can provide separate signals or combined or aggregated signals. In some examples, the sensors 152 can include a single sensor rather than providing separate sensors for each of the grain cleaning subsystems 118.

[0043] The separator loss sensors 148 provide signals indicative of grain loss in the left and right separators (not shown separately in FIG. 1). The separator loss sensors 148 can be associated with the left and right separators and can provide separate grain loss signals or a combined or aggregated signal. In some cases, sensing grain loss in the separators can also be performed using various different types of sensors. Figure 1

[0044] The agricultural harvester 100 can also include other sensors and measurement mechanisms. For example, the agricultural harvester 100 can include one or more of the following sensors: a header height sensor that senses a height of the header 102 above the ground 111; a stability sensor that senses a vibration or bounce (and amplitude) of the agricultural harvester 100; a residue setting sensor configured to sense whether the agricultural harvester 100 is configured to chop residue, produce a windrow, etc.; a clean grain bin fan speed sensor to sense a fan 120 speed; a concave gap sensor that senses a gap between the rotor 112 and the concave 114; a threshing rotor speed sensor that senses a rotor speed of the rotor 112; a chaffer screen gap sensor that senses a size of openings in the chaffer screen 122; a screen mesh gap sensor that senses a size of openings in the screen mesh 124; a material other than grain (MOG) moisture sensor that senses a moisture level of MOG passing through the agricultural harvester 100; one or more machine setting sensors configured to sense various configurable settings of the agricultural harvester 100; a machine orientation sensor that senses an orientation of the agricultural harvester 100; and a crop property sensor that senses various different types of crop properties such as crop type, crop moisture, and other crop properties.

[0045] The crop property sensor can also be configured to sense properties of cut crop material as the crop material is processed by the agricultural harvester 100. For example, in some cases, the crop property sensor can sense: grain quality such as broken grain, MOG levels; grain constituents such as starch and protein; and a grain feed rate as the grain travels through the feedhouse 106, the clean grain elevator 130, or elsewhere in the agricultural harvester 100. The crop property sensor can also sense a feed rate of biomass through the feedhouse 106, the separator 116, or elsewhere in the agricultural harvester 100. The crop property sensor can also sense a feed rate through the elevator 130 or through other portions of the agricultural harvester 100 as a grain mass flow rate, or provide other output signals indicative of other sensed variables.

[0046] ​Before describing how the agricultural harvester 100 generates a functional predictive weed map and uses the functional predictive weed map for control, a brief description of some items on the agricultural harvester 100 and their operation will first be described. Figure 2 and the description of FIG. 3 describes receiving a general type of prior information map and combining information from the prior information map with georeferenced sensor signals generated by in-field sensors, where the sensor signals are indicative of characteristics in a field, such as characteristics of crops or weeds present in the field. The characteristics of the field can include, but are not limited to, characteristics of the field (such as slope, weed intensity, weed type, soil moisture, surface quality); characteristics of crop properties (such as crop height, crop moisture, crop density, crop status); characteristics of grain properties (such as grain moisture, grain size, grain test weight); and characteristics of machine performance (such as loss level, job quality, fuel consumption, and power utilization). Relationships between characteristic values obtained from the in-field sensor signals and prior information map values are identified, and the relationships are used to generate a new functional predictive map. The functional predictive map predicts values at different geographic locations in the field, and one or more of the values can be used to control a machine, such as one or more subsystems of an agricultural harvester. In some cases, the functional predictive map can be presented to a user, such as an operator of an agricultural work machine, which can be an agricultural harvester. The functional predictive map can be presented to the user in a visual manner (such as by a display), in a tactile manner, or in an audible manner. The user can interact with the functional predictive map to perform editing operations and other user interface operations. In some cases, the functional predictive map can be used to control an agricultural work machine (such as an agricultural harvester), presented to an operator or other user, and presented to an operator or user for operator or user interaction, one or more of.

[0047] In reference to Figure 2 and the description of FIG. 3 describes a general method, reference is made to Figure 4 and Figure 5 a more specific method for generating a functional predictive weed map is described, which can be presented to an operator or user, or used to control the agricultural harvester 100, or both. Again, although this discussion is directed to an agricultural harvester, and in particular a combine harvester, the scope of the present disclosure encompasses other types of agricultural harvesters or other agricultural work machines.

[0048] Figure 2 is a block diagram showing some portions of an example agricultural harvester 100. Figure 2The agricultural harvester 100 is shown to illustratively include one or more processors or servers 201, a data store 202, a geo-location sensor 204, a communication system 206, and one or more field sensors 208 that sense one or more agricultural properties of a field while the harvesting operation is in progress. Agricultural properties can include any property that can have an effect on the harvesting operation. Some examples of agricultural properties include properties of the harvesting machine, the field, the plants on the field, and the weather. Other types of agricultural properties are also included. The field sensors 208 generate values corresponding to the sensed properties. The agricultural harvester 100 also includes a predictive model or relationship generator (hereinafter collectively referred to as “predictive model generator 210”), a prediction map generator 212, a control zone generator 213, a control system 214, one or more controllable subsystems 216, and an operator interface mechanism 218. The agricultural harvester 100 can also include a variety of other agricultural harvester functions 220. The field sensors 208 include, for example, on-board sensors 222, remote sensors 224, and other sensors 226 that sense properties of the field during the course of the agricultural operation. The predictive model generator 210 illustratively includes a prior information variable to field variable model generator 228, and the predictive model generator 210 can include other items 230. The control system 214 includes a communication system controller 229, an operator interface controller 231, a settings controller 232, a path planning controller 234, a feed rate controller 236, a header and reel controller 238, a belt conveyor belt controller 240, a table deck position controller 242, a residue system controller 244, a machine cleanout controller 245, a zone controller 247, and the system 214 can include other items 246. The controllable subsystems 216 include machine and header actuators 248, a propulsion subsystem 250, a steering subsystem 252, a residue subsystem 138, a machine cleanout subsystem 254, and the subsystems 216 can include a variety of other subsystems 256.

[0049] Figure 2 The agricultural harvester 100 is also shown to receive a prior information map 258. As described below, the prior information map 258 includes, for example, a vegetation index map or a vegetation map from a prior operation or a predicted weed map. However, the prior information map 258 can also encompass other types of data obtained prior to the harvesting operation or maps from prior operations. Figure 2An operator 260 is also shown as operating the agricultural harvester 100. The operator 260 interacts with the operator interface mechanism 218. In some examples, the operator interface mechanism 218 can include joysticks, levers, steering wheels, linkages, pedals, buttons, dials, keypads, user-actuatable elements on a user interface display device (such as icons, buttons, etc.), microphones and speakers (where voice recognition and speech synthesis are provided), and various other types of control devices. Where a touch-sensitive display system is provided, the operator 260 can interact with the operator interface mechanism 218 using touch gestures. These examples are provided as illustrative examples and are not intended to limit the scope of the present disclosure. Thus, other types of operator interface mechanisms 218 can be used and are within the scope of the present disclosure.

[0050] The prior information map 258 can be downloaded onto the agricultural harvester 100 and stored in the data storage 202 using the communication system 206 or otherwise. In some examples, the communication system 206 can be a cellular communication system, a system for communicating over a wide area network or a local area network, a system for communicating over a near field communication network, or a communication system configured to communicate over any of a variety of other networks or combinations of networks. The communication system 206 can also include a system that facilitates downloading or transferring information to and from a secure digital (SD) card or a universal serial bus (USB) card or both.

[0051] The geo-location sensor 204 illustratively senses or detects a geo-location or position of the agricultural harvester 100. The geo-location sensor 204 can include, without limitation, a global navigation satellite system (GNSS) receiver that receives signals from GNSS satellite transmitters. The geo-location sensor 204 can also include a real-time kinematic (RTK) component configured to improve the accuracy of position data derived from GNSS signals. The geo-location sensor 204 can include a dead reckoning system, a cellular triangulation system, or any of a variety of other geo-location sensors.

[0052] The field sensors 208 can be any of the field sensors described above with reference to the field sensor 208 of the agricultural harvester 100 of FIG. 1. Figure 1Any of the sensors described. The on-site sensors 208 include on-board sensors 222 mounted on the on-board agricultural harvester 100. Such sensors can include, for example, perception sensors (e.g., forward-looking monocular or stereo camera systems and image processing systems), image sensors inside the agricultural harvester 100 such as one or more clean grain cameras mounted to identify weed seeds exiting the agricultural harvester 100 through a residue system or from a clean grain system. The on-site sensors 208 also include remote on-site sensors 224 that capture on-site information. On-site data includes data acquired from sensors mounted on the harvester or data acquired by any sensor that detects data during the harvesting operation.

[0053] The predictive model generator 210 generates a model indicative of a relationship between values sensed by the field sensors 208 and a metric mapped to the field by the prior information map 258. For example, if the prior information map 258 maps vegetation index values to different locations in the field, and the field sensors 208 sense values indicative of weed intensity, the prior information variable to field variable model generator 228 generates a predictive weed model that models a relationship between vegetation index values and weed intensity values. The predictive weed model can also be generated based on vegetation index values from the prior information map 258 and multiple field data values generated by the field sensors 208. The predictive map generator 212 then generates a functional predictive weed map using the predictive weed model generated by the predictive model generator 210 that predicts values of a weed property (such as intensity) sensed by the field sensors 208 at different locations in the field based on the prior information map 258. In some examples, the type of values in the functional predictive map 263 can be the same as the type of field data sensed by the field sensors 208. In some cases, the type of values in the functional predictive map 263 can have different units than the data sensed by the field sensors 208. In some examples, the type of values in the functional predictive map 263 can be different than the type of data sensed by the field sensors 208, but related to the type of data sensed by the field sensors 208. For example, in some examples, the type of data sensed by the field sensors 208 can dictate the type of values in the functional predictive map 263. In some examples, the type of data in the functional predictive map 263 can be different than the type of data in the prior information map 258. In some cases, the type of data in the functional predictive map 263 can have different units than the data in the prior information map 258. In some examples, the type of data in the functional predictive map 263 can be different than the type of data in the prior information map 258, but related to the type of data in the prior information map 258. For example, in some examples, the type of data in the prior information map 258 can dictate the type of data in the functional predictive map 263. In some examples, the type of data in the functional predictive map 263 is different than one or both of the type of field data sensed by the field sensors 208 and the type of data in the prior information map 258. In some examples, the type of data in the functional predictive map 263 is the same as one or both of the type of field data sensed by the field sensors 208 and the type of data in the prior information map 258. In some examples, the type of data in the functional predictive map 263 is the same as one of the type of field data sensed by the field sensors 208 or the type of data in the prior information map 258, and different than the other.

[0054] Continuing with the previous example, where prior information map 258 is a vegetation index map and field sensor 208 senses values ​​indicating weed intensity, prediction map generator 212 can use the vegetation index values ​​from prior information map 258 and the model generated by prediction model generator 210 to generate a functional prediction map 263 predicting weed intensity at different locations in the field. Prediction map generator 212 then outputs prediction map 264.

[0055] like Figure 2 As shown, prediction map 264 uses prior information values ​​from prior information map 258 at various locations on the field and a prediction model to predict the values ​​of sensed characteristics (sensed by field sensors 208) or characteristics related to sensed characteristics at these locations. For example, if prediction model generator 210 has generated a prediction model indicating the relationship between vegetation index values ​​and weed intensity, then, given vegetation index values ​​at different locations on the field, prediction map generator 212 generates prediction map 264 predicting weed intensity values ​​at different locations on the field. The vegetation index values ​​at these locations obtained from the vegetation index map and the relationship between vegetation index values ​​and weed intensity obtained from the prediction model are used to generate prediction map 264.

[0056] The following will describe some changes in the data types mapped in prior information graph 258, the data types sensed by field sensor 208, and the data types predicted in prediction graph 264.

[0057] In some examples, the data type in the prior infographic 258 differs from the data type sensed by the field sensor 208, but the data type in the prediction infographic 264 is the same as the data type sensed by the field sensor 208. For example, the prior infographic 258 could be a vegetation index map, and the variable sensed by the field sensor 208 could be yield. The prediction infographic 264 could then be a predicted yield map that maps predicted yield values ​​to different geographic locations in the field. In another example, the prior infographic 258 could be a vegetation index map, and the variable sensed by the field sensor 208 could be crop height. The prediction infographic 264 could then be a predicted crop height map that maps predicted crop height values ​​to different geographic locations in the field.

[0058] Further, in some examples, the data type in the prior information map 258 is different from the data type sensed by the field sensor 208, and the data type in the prediction map 264 is different from both the data type in the prior information map 258 and the data type sensed by the field sensor 208. For example, the prior information map 258 can be a vegetation index map, and the variable sensed by the field sensor 208 can be crop height. The prediction map 264 can then be a predicted biomass map that maps predicted biomass values to different geographic locations in the field. In another example, the prior information map 258 can be a vegetation index map, and the variable sensed by the field sensor 208 can be yield. The prediction map 264 can then be a predicted speed map that maps predicted harvester speed values to different geographic locations in the field.

[0059] In some examples, the prior information map 258 is from a previous pass through the field during a prior operation, and the data type is different from the data type sensed by the field sensor 208, but the data type in the prediction map 264 is the same as the data type sensed by the field sensor 208. For example, the prior information map 258 can be a seed population map generated during planting, and the variable sensed by the field sensor 208 can be stem size. The prediction map 264 can then be a predicted stem size map that maps predicted stem size values to different geographic locations in the field. In another example, the prior information map 258 can be a seed mix map, and the variable sensed by the field sensor 208 can be crop status, such as standing crop or lodged crop. The prediction map 264 can then be a predicted crop status map that maps predicted crop status values to different geographic locations in the field.

[0060] In some examples, the prior information map 258 is from a previous pass through the field during a prior operation, and the data type is the same as the data type sensed by the field sensor 208, and the data type in the prediction map 264 is also the same as the data type sensed by the field sensor 208. For example, the prior information map 258 can be a yield map generated during the previous year, and the variable sensed by the field sensor 208 can be yield. The prediction map 264 can then be a predicted yield map that maps predicted yield values to different geographic locations in the field. In such an example, the prediction model generator 210 can use relative yield differences in the georeferenced prior information map 258 from the previous year to generate a prediction model that models a relationship between relative yield differences on the prior information map 258 and yield values sensed by the field sensor 208 during the current harvesting operation. The prediction model is then used by the prediction map generator 210 to generate the predicted yield map.

[0061] In another example, the prior information map 258 can be a weed intensity map generated during a prior operation, such as from a sprayer, and the variable sensed by the field sensor 208 can be weed intensity. The prediction map 264 can then be a predicted weed intensity map that maps predicted weed intensity values to different geographic locations in the field. In such an example, a map of weed intensity at the time of spraying is recorded in a georeferenced manner and provided to the agricultural harvester 100 as the prior information map 258 of weed intensity. The field sensor 208 can detect the weed intensity at the geographic location in the field at the time of harvesting, and the prediction model generator 210 can then establish a prediction model that models the relationship between the weed intensity at the time of harvesting and the weed intensity at the time of spraying. This is because the sprayer will have affected the weed intensity at the time of spraying, but the weeds can still reappear in similar areas by the time of harvesting. However, based on the time of harvesting, weather, weed type, etc., the weed areas at the time of harvesting can have different intensities.

[0062] In some examples, the prediction map 264 can be provided to the control zone generator 213. The control zone generator 213 groups adjacent portions of the field into one or more control zones based on the data values of the prediction map 264 associated with those adjacent portions. A control zone can include two or more contiguous portions of a field, such as a field, for which the control parameters corresponding to the control zone for controlling the controllable subsystem are constant. For example, the response time to change the settings of the controllable subsystem 216 can not be satisfactory to respond to changes in values contained in a map, such as the prediction map 264. In this case, the control zone generator 213 parses the map and identifies control zones of a defined size to accommodate the response time of the controllable subsystem 216. In another example, the size of the control zone can be determined to reduce wear caused by excessive actuator movement resulting from continuous adjustment. In some examples, there can be different control zone groups for each controllable subsystem 216 or group of controllable subsystems 216. The control zones can be added to the prediction map 264 to obtain a prediction control zone map 265. The prediction control zone map 265 can thus be similar to the prediction map 264, except that the prediction control zone map 265 includes control zone information defining the control zones. Thus, as described herein, a functional prediction map 263 can or can not include control zones. Both the prediction map 264 and the prediction control zone map 265 are functional prediction maps 263. In one example, the functional prediction map 263 does not include control zones, such as the prediction map 264. In another example, the functional prediction map 263 does include control zones, such as the prediction control zone map 265. In some examples, if an intercropping production system is implemented, there can be multiple crops in the field at the same time. In this case, the prediction map generator 212 and the control zone generator 213 are able to identify the location and characteristics of the two or more crops and then generate the prediction map 264 and the prediction map with control zones 265 accordingly.

[0063] It should also be appreciated that the control zone generator 213 can cluster values to generate control zones, and the control zones can be added to the predicted control zone map 265 or a separate map, showing only the generated control zones. In some examples, the control zones can be used to control or calibrate the agricultural harvester 100 or both. In other examples, the control zones can be presented to the operator 260 and used to control or calibrate the agricultural harvester 100, and in other examples, the control zones can be presented to the operator 260 or another user or stored for later use.

[0064] The predicted map 264 or the predicted control zone map 265 or both are provided to a control system 214 that generates control signals based on the predicted map 264 or the predicted control zone map 265 or both. In some examples, the communication system controller 229 controls the communication system 206 to communicate the predicted map 264 or the predicted control zone map 265 or control signals based on the predicted map 264 or the predicted control zone map 265 to other agricultural harvesters that are harvesting in the same field. In some examples, the communication system controller 229 controls the communication system 206 to send the predicted map 264, the predicted control zone map 265 or both to other remote systems.

[0065] The operator interface controller 231 is operable to generate control signals to control the operator interface mechanisms 218. The operator interface controller 231 is also operable to present the prediction map 264 or the prediction control zone map 265 or other information derived from or based on the prediction map 264, the prediction control zone map 265, or both, to the operator 260. The operator 260 can be a local operator or a remote operator. As an example, the controller 231 generates control signals to control the display mechanisms to display one or both of the prediction map 264 and the prediction control zone map 265 to the operator 260. The controller 231 can generate operator actuatable mechanisms, which are shown and can be actuated by the operator to interact with the displayed maps. The operator can edit the maps by, for example, correcting the displayed weed types on the map based on the operator’s observations. The settings controller 232 can generate control signals to control various settings on the agricultural harvester 100 based on the prediction map 264, the prediction control zone map 265, or both. For example, the settings controller 232 can generate control signals to control the machine and header actuators 248. In response to the generated control signals, the machine and header actuators 248 operate to control, for example, one or more of the sieve and chaffer settings, the threshing machine gap, the rotor settings, the clean grain fan speed settings, the header height, the header functions, the reel speed, the reel position, the belt conveyor functions (where the agricultural harvester 100 is coupled to a belt conveyor header), the corn header functions, the in-cab distribution control, and other actuators 248 that affect other functions of the agricultural harvester 100. The path planning controller 234 illustratively generates control signals to control the steering subsystem 252 to steer the agricultural harvester 100 according to a desired path. The path planning controller 234 can control the path planning system to generate a route for the agricultural harvester 100 and can control the propulsion subsystem 250 and the steering subsystem 252 to steer the agricultural harvester 100 along the route. The feed rate controller 236 can control various subsystems, such as the propulsion subsystem 250 and the machine actuators 248, to control the feed rate based on the prediction map 264 or the prediction control zone map 265 or both. For example, the feed rate controller 236 can reduce the speed of the machine 100 as the agricultural harvester 100 approaches a patch of weeds with a high intensity value above a selected threshold to maintain a constant feed rate of the biomass through the machine. The header and reel controller 238 can generate control signals to control the header or the reel or other header functions. The belt conveyor belt controller 240 can generate control signals to control the belt conveyor belt or other belt conveyor functions based on the prediction map 264, the prediction control zone map 265, or both.The header position controller 242 can generate control signals to control the position of the header deck included on the header based on the prediction map 264 or the prediction control zone map 265 or both, and the residue system controller 244 can generate control signals to control the residue subsystem 138 based on the prediction map 264 or the prediction control zone map 265 or both. The machine cleanout controller 245 can generate control signals to control the machine cleanout subsystem 254. For example, based on the different types of seeds or weeds passing through the machine 100, a particular type of machine cleanout operation or frequency of performing a cleanout operation can be controlled. Other controllers included on the agricultural harvester 100 can also control other subsystems based on the prediction map 264 or the prediction control zone map 265 or both.

[0066] Figure 3A and Figure 3B FIG. 3 (collectively referred to herein as FIG. 3) illustrates a flowchart diagram illustrating one example of operations of the agricultural harvester 100 in generating the prediction map 264 and the prediction control zone map 265 based on the prior information map 258.

[0067] At 280, the agricultural harvester 100 receives the prior information map 258. Examples of the prior information map 258 or receiving the prior information map 258 are discussed with respect to blocks 281, 282, 284, and 286. As discussed above, the prior information map 258 maps values of a variable corresponding to a first characteristic to different locations in the field, as shown in block 282. As shown in block 281, receiving the prior information map 258 can include selecting one or more of a plurality of possible prior information maps available. For example, one prior information map can be a vegetation index map generated from aerial images. Another prior information map can be a map generated during a previous pass through the field, which can be performed by a different machine in the field, such as a sprayer or other machine, performing a previous operation. The process of selecting one or more prior information maps can be manual, semi-automatic, or automatic. The prior information map 258 is based on data collected prior to the current harvesting operation. This is indicated by block 284. For example, the data can be collected based on aerial images taken during a previous year or early in the current growing season or other time. The data can be based on data detected in a manner other than using aerial images. For example, the agricultural harvester 100 can be equipped with a sensor, such as an internal optical sensor, that identifies weed seeds exiting the agricultural harvester 100. The weed seed data detected by the sensor during a previous year of harvesting can be used as data for generating the prior information map 258. The sensed weed data can be combined with other data to generate the prior information map 258. For example, based on the number of weed seeds exiting the agricultural harvester 100 at different locations and based on other factors, such as whether the seeds were spread by a spreader or fell into a pile of material, weather conditions (such as wind) at the time the seeds fell or were spread, drainage conditions that can move the seeds around the field, or other information, the locations of these weed seeds can be predicted, such that the prior information map 258 maps the predicted seed locations in the field. The data of the prior information map 258 can be transmitted to the agricultural harvester 100 using the communication system 206 and stored in the data storage device 202. The data of the prior information map 258 can also be provided to the agricultural harvester 100 in other ways using the communication system 206, and this is represented by block 286 in the flowchart of FIG. 3. In some examples, the prior information map 258 can be received by the communication system 206.

[0068] At the start of a harvesting operation, the field sensors 208 generate sensor signals indicative of one or more field data values indicative of a characteristic, e.g., a plant characteristic, such as a weed characteristic, as indicated by block 288. Examples of field sensors 288 are discussed with respect to blocks 222, 290, and 226. As explained above, the field sensors 208 include on-board sensors 222, remote field sensors 224, such as UAV-based sensors that gather field data on each flight (as shown in block 290), or other types of field sensors specified by the field sensors 226. In some examples, data from the on-board sensors is georeferenced using position, heading, or velocity data from the geolocation sensors 204.

[0069] The predictive model generator 210 controls the prior information variables to the field variables model generator 228 to generate a model that models the relationship between the mapped values contained in the prior information map 258 and the field values sensed by the field sensors 208, as indicated by block 292. The characteristic or data type represented by the mapped values in the prior information map 258 and the field values sensed by the field sensors 208 can be the same characteristic or data type or different characteristics or data types.

[0070] The relationship or model generated by the predictive model generator 210 is provided to the predictive map generator 212. The predictive map generator 212 uses the predictive model and the prior information map 258 to generate a predictive map 264 that predicts values of the characteristic sensed by the field sensors 208 at different geographic locations in the field being harvested, or a different characteristic related to the characteristic sensed by the field sensors 208, as indicated by block 294.

[0071] It should be noted that in some examples, the prior information map 258 can include two or more different maps or two or more different layers of a single map. Each layer can represent a different data type than the data type of another layer, or the layers can have the same data type obtained at different times. Each map of the two or more different maps or each layer of the two or more different layers of a map maps a different type of variable to a geographic location in the field. In such examples, the prediction model generator 210 generates a prediction model that models a relationship between the on-site data and each of the different variables mapped by the two or more different maps or the two or more different layers of a map. Similarly, the on-site sensors 208 can include two or more sensors, each of which senses a different type of variable. Thus, the prediction model generator 210 generates a prediction model that models a relationship between each type of variable mapped by the prior information map 258 and each type of variable sensed by the on-site sensors 208. The prediction map generator 212 can use the prediction model and each of the maps or layers of the map in the prior information map 258 to generate a functional prediction map 263 that predicts values of each sensed characteristic (or a characteristic related to the sensed characteristic) sensed by the on-site sensors 208 at different locations in the field being harvested.

[0072] The prediction map generator 212 configures the prediction map 264 such that the prediction map 264 is operable (or consumable) by the control system 214. The prediction map generator 212 can provide the prediction map 264 to the control system 214 or to the control zone generator 213 or to both. Some examples of different ways in which the prediction map 264 can be configured or output are described with respect to blocks 296, 295, 299, and 297. For example, the prediction map generator 212 configures the prediction map 264 such that the prediction map 264 includes values that can be read by the control system 214 and used as a basis for generating control signals for one or more of the different controllable subsystems of the agricultural harvester 100, as indicated by block 296.

[0073] The control zone generator 213 can divide the prediction map 264 into control zones based on the values on the prediction map 264. Contiguously geolocated values within a threshold of each other can be grouped into control zones. The threshold can be a default threshold, or the threshold can be set based on operator input, based on input from an automation system, or based on other criteria. The size of the zones can be based on the responsiveness of the control system 214, controllable subsystems 216, based on wear considerations, or based on other criteria, as indicated by block 295. The prediction map generator 212 configures the prediction map 264 for presentation to an operator or other user. The control zone generator 213 can configure the prediction control zone map 265 for presentation to an operator or other user. This is indicated by block 299. When presented to an operator or other user, the presentation of the prediction map 264 or the prediction control zone map 265 or both can include one or more of the predicted values on the prediction map 264 related to geographic locations, the control zones on the prediction control zone map 265 related to geographic locations, and the set values or control parameters used based on the predicted values on the map 264 or the zones on the prediction control zone map 265. In another example, the presentation can include more abstract information or more detailed information. The presentation can also include a confidence level that indicates the accuracy of the predicted values on the prediction map 264 or the zones on the prediction control zone map 265 to match measured values that can be measured by sensors on the agricultural harvester 100 as the agricultural harvester 100 moves through the field. Additionally, where information is presented to more than one location, an authentication and authorization system can be provided to implement authentication and authorization processes. For example, there can be a hierarchy of individuals that are authorized to view and change the maps and other presented information. As an example, an onboard display device can display the maps locally on the machine in near real time, or the maps can also be generated at one or more remote locations, or both. In some examples, each physical display device at each location can be associated with a person or user permission level. The user permission level can be used to determine which display indicia are visible on the physical display device, and which values the corresponding person can change. For example, a local operator of the machine 100 can not be able to see the information corresponding to the prediction map 264 or make any changes to the machine operation. However, a supervisor such as a supervisor at a remote location can be able to see the prediction map 264 on a display, but be prevented from making any changes. A manager that can be at a separate remote location can be able to see all of the elements on the prediction map 264 and also be able to change the prediction map 264. In some cases, the prediction map 264 can be accessible and changeable by a manager located remotely, can be used for machine control. This is one example of an authorization hierarchy that can be implemented. The prediction map 264 or the prediction control zone map 265 or both can also be configured in other ways, as indicated by block 297.

[0074] At block 298, inputs from the geo-location sensor 204 and other field sensors 208 are received by the control system. In particular, at block 300, the control system 214 detects input from the geo-location sensor 204 identifying the geo-location of the agricultural harvester 100. Block 302 represents sensor input indicating the trajectory or heading of the agricultural harvester 100 being received by the control system 214, and block 304 represents the speed of the agricultural harvester 100 being received by the control system 214. Block 306 represents other information being received by the control system 214 from the various field sensors 208.

[0075] At block 308, the control system 214 generates control signals to control the controllable subsystems 216 based on the prediction map 264 or the prediction control zone map 265 or both, as well as the inputs from the geo-location sensor 204 and any other field sensors 208. At block 310, the control system 214 applies the control signals to the controllable subsystems. It will be appreciated that the particular control signals that are generated and the particular controllable subsystems 216 that are controlled can vary based on one or more different factors. For example, the control signals that are generated and the controllable subsystems 216 that are controlled can be based on the type of prediction map 264 or prediction control zone map 265 or both that is being used. Similarly, the control signals that are generated, the controllable subsystems 216 that are controlled, and the timing of the control signals can be based on various delays in the flow of the crop through the agricultural harvester 100 and the responsiveness of the controllable subsystems 216.

[0076] As an example, the predicted map 264 generated in the form of a predicted weed map can be used to control one or more of the subsystems 216. For example, the predicted weed map can include weed intensity values that geographically reference locations within the field being harvested. The weed intensity values from the predicted weed map can be extracted and used to control the steering and propulsion subsystems 252 and 250. By controlling the steering and propulsion subsystems 252 and 250, the feed rate of material moving through the agricultural harvester 100 can be controlled. Similarly, the header height can be controlled to take more or less material, and thus, the header height can also be controlled to control the feed rate of material through the agricultural harvester 100. In other examples, control of the header height can be implemented if the predicted map 264 maps weed height relative to locations in the field. For example, if the values present in the predicted weed map indicate that one or more areas have weed heights in a first amount of height, the header and reel controller 238 can control the header height such that the header is positioned above the first amount of height of weeds when performing a harvesting operation in the one or more areas having weeds in the first amount of height. Thus, the header and reel controller 238 can be controlled using the geographically referenced values present in the predicted weed map to position the header to a height above the predicted height values of weeds obtained from the predicted weed map. Additionally, the header height can be automatically changed by the header and reel controller 238 as the agricultural harvester 100 progresses through the field using the geographically referenced values obtained from the predicted weed map. The preceding examples involving the use of weed height and intensity from the predicted weed map are provided by way of example only. Thus, a variety of other control signals can be generated using values obtained from the predicted weed map or other types of predicted maps to control one or more of the controllable subsystems 216.

[0077] At block 312, a determination is made as to whether the harvesting operation has been completed. If harvesting has not been completed, the process proceeds to block 314 where the reading of the field sensor data from the geo-location sensors 204 and the field sensors 208 (and possibly other sensors) continues.

[0078] In some examples, at block 316, the agricultural harvester 100 can also detect learning trigger criteria to perform machine learning on one or more of the predicted map 264, the predicted control zone map 265, the models generated by the predicted model generator 210, the zones generated by the control zone generator 213, the one or more control algorithms implemented by the controllers in the control system 214, and other triggers for learning.

[0079] The learning trigger criteria can include any of a variety of different criteria. Some examples of detecting trigger criteria are discussed with respect to blocks 318, 320, 321, 322, and 324. For example, in some examples, triggering learning can include recreating the relationships used to generate the prediction model when a threshold amount of field sensor data is obtained from the field sensors 208. In such examples, the amount of field sensor data received from the field sensors 208 that exceeds the threshold triggers or causes the prediction model generator 210 to generate a new prediction model used by the prediction map generator 212. Thus, as the agricultural harvester 100 continues harvesting operations, the threshold amount of field sensor data received from the field sensors 208 triggers the creation of new relationships represented by the prediction models generated by the prediction model generator 210. Further, the new prediction model can be used to regenerate a new prediction map 264, a prediction control zone map 265, or both. Block 318 represents detecting the threshold amount of field sensor data used to trigger the creation of a new prediction model.

[0080] In other examples, the learning trigger criteria can be based on the degree of change in the field sensor data from the field sensors 208, such as the degree of change over time or compared to previous values. For example, if the change within the field sensor data (or the relationship between the field sensor data and information in the priori information map 258) is within a selected range, or less than a defined amount, or below a threshold, then a new prediction model is not generated by the prediction model generator 210. As a result, the prediction map generator 212 does not generate a new prediction map 264, a prediction control zone map 265, or both. However, if the change within the field sensor data is outside of the selected range, greater than the defined amount, or above the threshold, for example, then the prediction model generator 210 generates a new prediction model using all or a portion of the newly received field sensor data used by the prediction map generator 212 to generate a new prediction map 264. At block 320, the change in the field sensor data, such as the size of the amount of data that exceeds the selected range or the size of the change in the relationship between the field sensor data and information in the priori information map 258, can be used as a trigger that results in the generation of a new prediction model and prediction map. Continuing with the example described above, the threshold, range, and defined amount can be set to a default value, set by an operator or user through interaction with a user interface, set by an automated system, or otherwise set.

[0081] Other learning trigger criteria can also be used. For example, if the prediction model generator 210 switches to a different prior information map (different from the initially selected prior information map 258), the switch to the different prior information map can trigger relearning by the prediction model generator 210, the prediction map generator 212, the control zone generator 213, the control system 214, or other. In another example, a transition of the agricultural harvester 100 to a different terrain or to a different control zone can also be used as a learning trigger criterion.

[0082] In some cases, the operator 260 can also edit the prediction map 264 or the prediction control zone map 265 or both. The editing can change values on the prediction map 264, change the size, shape, location, or existence of control zones on the prediction control zone map 265, or both. Block 321 shows that the edited information can be used as a learning trigger criterion.

[0083] In certain cases, the operator 260 can also observe that the automatic control of the controllable subsystem is not as the operator 260 desires. In this case, the operator 260 can provide a manual adjustment to the controllable subsystem, which reflects that the operator 260 desires the controllable subsystem to operate differently than commanded by the control system 214. Thus, the manual change to the setting by the operator 260 can cause one or more of the following: the prediction model generator 210 to relearn a model, the prediction map generator 212 to regenerate the map 264, the control zone generator 213 to regenerate one or more control zones on the prediction control zone map 265, and the control system 214 to relearn a control algorithm or perform machine learning on one or more of the controller components 232-246 in the control system 214, based on the adjustment by the operator 260, as shown in block 322. Block 324 represents using other trigger learning criteria.

[0084] In other examples, relearning can be performed periodically or intermittently, for example based on a selected time interval, such as a discrete time interval or a variable time interval, as shown in block 326.

[0085] If relearning is triggered (whether based on a learning trigger criterion or based on the passage of a time interval, as shown in block 326), one or more of the prediction model generator 210, the prediction map generator 212, the control zone generator 213, and the control system 214 perform machine learning to generate a new prediction model, a new prediction map, new control zones, and a new control algorithm, respectively, based on the learning trigger criterion. The new prediction model, the new prediction map, and the new control algorithm are generated using any additional data collected since the last performance of a learning operation. The performance of the relearning is indicated by block 328.

[0086] If the harvesting operation has been completed, the operation moves from block 312 to block 330, in which one or more of the prediction map 264, the prediction control zone map 265, and the prediction model generated by the prediction model generator 210 are stored. The prediction map 264, the prediction control zone map 265, and the prediction model can be stored locally on the data storage 202 or transmitted to a remote system for later use using the communication system 206.

[0087] It will be noted that while some examples herein describe the prediction model generator 210 and the prediction map generator 212 receiving a prior information map when generating a prediction model and a functional prediction map, respectively, in other examples, the prediction model generator 210 and the prediction map generator 212 can receive other types of maps, including prediction maps, such as a functional prediction map generated during a harvesting operation.

[0088] Figure 4 is Figure 1 A block diagram of a portion of the agricultural harvester 100 shown in FIG. 1. In particular, Figure 4 Examples of the prediction model generator 210 and the prediction map generator 212 are shown in particular more detail. Figure 4 The information flow between the different components shown is also illustrated. The prediction model generator 210 receives a vegetation index map 332 as a prior information map. The prediction model generator 210 also receives a geographic location 334 or an indication of a geographic location from the geographic location sensor 204. The on-site sensors 208 schematically include a weed sensor, such as a weed sensor 336, as well as a processing system 338. In some cases, the weed sensor 336 can be located on the agricultural harvester 100. The processing system 338 processes sensor data generated from the on-board weed sensor 336 to generate processed data, some examples of which are described below.

[0089] In some examples, the weed sensor 336 can be an optical sensor, such as a camera, that generates an image of an area of the field to be harvested. In some cases, the optical sensor can be arranged on the agricultural harvester 100 to collect images of an area adjacent to the agricultural harvester 100, such as an area located in front of, to the side of, behind, or in another direction relative to the agricultural harvester 100 as the agricultural harvester 100 moves through the field during a harvesting operation. The optical sensor can also be located on or inside the agricultural harvester 100 to obtain images of one or more portions outside or inside the agricultural harvester 100. The processing system 338 processes the image or images obtained by the weed sensor 336 to generate processed image data that identifies one or more characteristics of weeds in the image. The weed characteristics detected by the processing system 338 can include the location of weeds present in the image, the intensity of weed patches in the image, or the type of weeds in the image.

[0090] The field sensor 208 can be or include other types of sensors, such as a camera positioned along a path of travel of cut crop material in the agricultural harvester 100 (hereinafter referred to as a “process camera”). The process camera can be located inside the agricultural harvester 100 and can capture images of the crop material (including seeds) as the crop material moves through or is expelled from the agricultural harvester 100. The process camera can obtain images of the seeds, and the image processing system 338 is operable to identify the presence of weed seeds, detect an amount of weed seeds detected (such as a number of weed seeds (to give an indication of a density of weeds encountered in a field)), and identify one or more weed types based on the types of seeds identified in the images. Thus, in some examples, the processing system 338 is operable to detect the presence of weed seeds in cut crop material passing through the agricultural harvester 100, an amount of weed seeds present in the cut crop material (e.g., an amount per volume of the cut crop material), and weed types corresponding to the detected weed seeds encountered by the agricultural harvester 100 during the course of a harvesting operation.

[0091] In other examples, the weed sensor 336 can rely on any wavelength(s) of electromagnetic energy and the way in which the electromagnetic energy is reflected, absorbed, attenuated, or transmitted through weed seeds or biomass. The weed sensor 336 can sense other electromagnetic properties of weed seeds and biomass, such as a dielectric constant, as the cut crop material passes between two capacitive plates. The weed sensor 336 can also rely on machine properties of the seeds and biomass, such as a signal generated when a weed seed impacts a piezoelectric sheet or when an impact by a seed is detected by a microphone or accelerometer. Other material properties and sensors can also be used. In some examples, raw data or processed data from the weed sensor 336 can be presented to the operator 260 via the operator interface mechanism 218. The operator 260 can be on the agricultural harvester 100 or at a remote location.

[0092] The present discussion is directed to examples in which the weed sensor 336 is an image sensor, such as a camera. It should be understood that this is merely one example, and that the above-mentioned sensors are also contemplated herein as other examples of the weed sensor 336. As Figure 4 As shown, the example prediction model generator 210 includes one or more of a weed presence to vegetation index model generator 342, a weed intensity to vegetation index model generator 344, and a weed type to vegetation index model generator 346. In other examples, the vegetation index model generator 210 can include more or fewer of these components, or other components. Figure 4The prediction model generator 210 can include additional components, fewer components, or different components than those shown in the example of FIG. 2. Thus, in some examples, the prediction model generator 210 can also include other items 348, which can include other types of prediction model generators to generate other types of weed characteristic models.

[0093] The model generator 342 identifies a relationship between the presence of weeds detected in the image data 340 at a geographic location corresponding to where the image data 340 was obtained and the vegetation index values from the vegetation index map 332 corresponding to the same locations in the field where the weed characteristics were detected. Based on this relationship established by the model generator 342, the model generator 342 generates a predictive weed model. The predictive weed model is used by the weed location map generator 352 to predict the presence of weeds at different locations in the field based on the georeferenced vegetation index values contained in the vegetation index map 332 at the same locations in the field.

[0094] The model generator 344 identifies a relationship between the weed intensity levels represented in the processed image data 340 at a geographic location corresponding to the image data 340 and the vegetation index values at the same geographic location. Again, the vegetation index values are georeferenced values contained in the vegetation index map 332. The model generator 344 then generates a predictive weed model that is used by the weed intensity map generator 354 to predict the weed intensity at a location in the field based on the vegetation index value for that location.

[0095] The model generator 346 identifies a relationship between the weed type at a particular location in the field identified by the processed image data 340 and the vegetation index values from the vegetation index map 332 at the same location. The model generator 346 generates a predictive weed model that is used by the weed type map generator 356 to predict the weed type at a particular location in the field based on the vegetation index value for that location.

[0096] In view of the above, the prediction model generator 210 is operable to generate a plurality of predictive weed models, such as one or more of the predictive weed models generated by the model generators 342, 344, and 346. In another example, two or more of the predictive weed models described above can be combined into a single predictive weed model that predicts two or more of weed location, weed intensity, and weed type based on vegetation index values at different locations in the field. Any one or combination of these weed models is represented by the weed model 350 in Figure 4 In some examples, the prediction model generator 210 can generate a single predictive weed model that predicts two or more of weed location, weed intensity, and weed type based on vegetation index values at different locations in the field. In this example, the weed model 350 represents the single predictive weed model.

[0097] The predictive weed model 350 is provided to the prediction map generator 212. In Figure 4In the example of FIG. 3, the prediction model generator 210 includes a weed location model 350, a weed intensity model 352, and a weed type model 354. In other examples, the prediction model generator 210 can include additional, fewer, or different models. Thus, in some examples, the prediction model generator 210 can include other items 356, which can include other types of models to generate predictions for other types of weed characteristics. The weed location model 350 receives the vegetation index values and the vegetation index map 332 and generates a prediction of the presence of weeds at different locations in the field based on the vegetation index values and the prediction model 350.

[0098] The weed intensity model 352 generates a prediction of the intensity of weeds at different locations in the field based on the vegetation index values and the prediction model 350. The weed type model 354 generates a prediction of the type of weeds at different locations in the field based on the vegetation index values and the prediction model 350.

[0099] The prediction model generator 210 outputs one or more prediction maps 360 of one or more of the predicted weed locations, weed intensities, or weed types. Each of the prediction maps 360 predicts a respective weed characteristic at different locations in the field. Each of the generated prediction maps 360 can be provided to the control zone generator 213, the control system 214, or both. The control zone generator 213 generates control zones and merges the control zones into a functional prediction map (i.e., the prediction maps 360) to produce a prediction control zone map 265. One or both of the prediction map 264 and the prediction control zone map 265 can be provided to the control system 214, which generates control signals to control one or more of the controllable subsystems 216 based on the prediction map 264, the prediction control zone map 265, or both.

[0100] Figure 5 is a flowchart of an example of the operations of the prediction model generator 210 and the prediction map generator 212 in generating the prediction weed model 350 and the prediction weed maps 360. At block 362, the prediction model generator 210 and the prediction map generator 212 receive the previous vegetation index map 332. At block 364, the processing system 338 receives one or more images from the weed sensor 336. As discussed above, the weed sensor 336 can be a camera, such as the forward-facing camera 366; an optical sensor 368 that observes the interior of the combine, such as a camera; or another type of on-board weed sensor 370.

[0101] At block 372, the processing system 338 processes the one or more received images to generate image data indicative of a characteristic of the weeds present in the one or more images. At block 374, the image data can be indicative of a weed location, a weed intensity, or both, that can be present at a location, such as a location in front of the combine harvester. In some cases, as shown in block 376, the image data can be indicative of weed seeds located inside the combine harvester or expelled from the combine harvester. In some cases, as shown in block 380, the image data can be indicative of a weed type. Thus, the image data includes a weed type indicator 378 that identifies a type of weed(s) encountered by the combine harvester. The weed type can be determined based on one or more images of weed plants, one or more images of weed seeds, or one or more images containing subject matter indicative of a weed type. The image data can also include other data.

[0102] At block 382, the predictive model generator 210 also obtains a geographic location corresponding to the image data. For example, the predictive model generator 210 can obtain the geographic location from the geographic location sensor 204 and determine the precise geographic location at which the image was taken or from which the image data 340 was derived based on machine delay, machine speed, and the like.

[0103] At block 384, the predictive model generator 210 generates one or more predictive weed models, such as the weed models 350, that model a relationship between vegetation index values obtained from a prior information map, such as the prior information map 258, and a weed characteristic or related characteristic sensed by the field sensor 208. For example, the predictive model generator 210 can generate a predictive weed model that models a relationship between vegetation index values and a sensed characteristic that includes a weed location, a weed intensity, or a weed type indicated by the image data obtained from the field sensor 208.

[0104] At block 386, the predictive weed models, such as the predictive weed models 350, are provided to the predictive map generator 212, which generates a predictive weed map 360 that maps a predicted weed characteristic based on the vegetation index map and the predictive weed models 350. For example, in some examples, the predictive weed map 360 predicts a weed location. In some examples, the predictive weed map 360 predicts a weed location as well as a weed intensity value, as shown in block 388. In some examples, the predictive weed map 360 predicts a weed location and a weed type, as shown in block 390, and in other examples, the predictive map 360 predicts other items, as shown in block 392. Further, the predictive weed map 360 can be generated during the course of an agricultural operation. Thus, as the agricultural harvester moves through a field in order to perform an agricultural operation, the predictive weed map 360 is generated as the agricultural operation is being performed.

[0105] In block 394, the predictive map generator 212 outputs a predicted weed map 360. In block 391, the predictive weed map generator 212 outputs the predicted weed map for presentation to operator 260 and possible interaction by operator 260. In block 393, the predictive map generator 212 can configure the map for use by control system 214. In block 395, the predictive map generator 212 can also provide map 360 to control area generator 213 to generate control areas. In block 397, the predictive map generator 212 also configures map 360 in other ways. The predicted weed map 360 (with or without control areas) is provided to control system 214. In block 396, control system 214 generates control signals based on the predicted weed map 360 to control controllable subsystem 216.

[0106] Figure 6A yes Figure 1 A block diagram of an example portion of an agricultural harvester 100 is shown. Specifically, Figure 6A Examples of the prediction model generator 210 and the prediction map generator 212 are shown in particular. In the illustrated example, the infographic is one or more of a vegetation index map 332, a predicted weed map 360, or a priori operation map 400. The priori operation map 400 may include vegetation values ​​(such as vegetation index values ​​or other vegetation values) indicating weed intensity or weed type at different locations in the field. The vegetation values ​​may be vegetation values ​​collected during a priori operation (such as a priori operation performed by a sprayer).

[0107] Moreover, in Figure 6A In the example shown, field sensor 208 may include one or more of biomass sensor 402, machine speed sensor 146, operator input sensor 404, and processing system 406. Field sensor 208 may also include other sensors 408.

[0108] The biomass sensor 402 senses a variable indicative of the biomass of the material being processed by the agricultural harvester 100. In some examples, the biomass sensor 402 can be an optical sensor 410, such as one of the optical sensors or cameras discussed above. In some examples, the biomass sensor 402 can be a rotor pressure sensor 412 or another sensor 414. The rotor pressure sensor 412 can sense a rotor drive pressure of the threshing rotor 112. The rotor drive pressure of the threshing rotor 112 is indicative of the torque being applied by the rotor 112 on the material being processed by the agricultural harvester 100. As the biomass of the material being processed by the agricultural harvester 100 increases, the rotor drive pressure also increases. Thus, by sensing the rotor drive pressure, an indication of the biomass of the material being processed can be obtained. The moisture sensor 403 senses a variable indicative of the moisture of the material being processed by or proximate to the agricultural harvester 100. The moisture sensor 403 can sense the material in the agricultural harvester 100 or the material proximate to the agricultural harvester 100. The moisture sensor 403 can include a capacitive sensor, a resistive sensor, or other sensor that can measure the moisture of the material. In some examples, the moisture sensor 403 senses an amount of moisture on the plants, such as dew or precipitation. The above discussion regarding the optical sensor 410 can apply to the moisture sensor 403. Figure 1 One example of a machine speed sensor 146 is discussed above. The machine speed sensor 146 senses a travel speed (e.g., ground speed) of the agricultural harvester 100 or a variable indicative of the travel speed of the agricultural harvester 100.

[0109] The operator input sensor 404 illustratively senses various operator inputs. The inputs can be setting inputs for controlling settings on the agricultural harvester 100 or other control inputs, such as steering inputs and other inputs. Thus, when the operator 260 changes a setting or provides a command input through the operator interface mechanism 218, such input is detected by the operator input sensor 404, which provides a sensor signal indicative of the sensed operator input. The processing system 406 can receive the sensor signal from the biomass sensor 402 or the operator input sensor 404 or both and generate an output indicative of the sensed variable. For example, the processing system 406 can receive the sensor input from the optical sensor 410 or the rotor pressure sensor 412 and generate an output indicative of the biomass. The processing system 406 can also receive input from the operator input sensor 404 and generate an output indicative of the sensed operator input.

[0110] The prediction model generator 210 can include a weed characteristic to biomass model generator 416, a weed characteristic to moisture model generator 417, a weed characteristic to speed model generator 420, and a weed characteristic to operator command model generator 422. In other examples, the prediction model generator 210 can include additional, fewer, or other model generators 434. The prediction model generator 210 can receive the geographic location indicator 334 from the geographic location sensor 204 and generate prediction models 422 that model relationships between information in one or more of the information maps and one or more of the following items: biomass sensed by the biomass sensor 402, machine speed sensed by the machine speed sensor 146, and operator input commands sensed by the operator input sensor 404. For example, the weed characteristic to biomass model generator 416 generates a relationship between a weed characteristic value (which can be on the vegetation index map 332, the predicted weed map 360, or the prior operation map 400) and a biomass value sensed by the biomass sensor 402. The weed characteristic to moisture model generator 417 illustratively generates a model representing a relationship between a weed characteristic and a variable indicative of moisture in or on plants in the field sensed by the moisture sensor 403. The weed characteristic to speed model generator 420 illustratively generates a model representing a relationship between a weed characteristic and a travel speed or a variable indicative of a travel speed sensed by the machine speed sensor 146. The weed characteristic to operator command model generator 422 generates a model that models a relationship between a weed characteristic reflected on the vegetation index map 332, the predicted weed map 360, the prior operation map 400, or any combination thereof and operator input commands sensed by the operator input sensor 404. The prediction models 426 generated by the prediction model generator 210 can include one or more of the prediction models that can be generated by the weed characteristic to biomass model generator 416, the weed characteristic to speed model generator 420, the weed characteristic to operator command model generator 422, and other model generators that can be included as part of other items 424.

[0111] In Figure 6AIn the example of FIG. 4, the prediction map generator 212 includes a predicted biomass map generator 428, a predicted moisture map generator 429, a predicted machine speed map generator 430, and a predicted operator command map generator 432. In other examples, the prediction map generator 212 can include additional, fewer, or other map generators 434. The predicted biomass map generator 428 receives a prediction model 426 modeling a relationship between weed characteristics and biomass (such as a prediction model generated by the weed characteristics to biomass model generator 416) and one or more of the information maps. The predicted biomass map generator 428 generates a functional predicted biomass map 436 of predicted biomass at different locations in the field based on one or more of the weed characteristics in one or more of the information maps at those locations in the field and based on the prediction model 426.

[0112] The predicted moisture map generator 429 receives a prediction model 426 modeling a relationship between weed characteristics and material moisture (such as a prediction model generated by the weed characteristics to moisture model generator 417) and generates a functional predicted moisture map 437 of predicted material moisture at different locations in the field based on vegetation index values in the vegetation index map 332, or predicted weed intensity values in the predicted weed map 360, or vegetation values in the prior operation map 400 at those locations in the field and based on the prediction model 426. Material moisture, especially weed plant moisture, can cause seed staining on edible grains. Thus, to avoid seed staining, for example, the predicted moisture map 437 can be used to avoid wet areas of the field until dew present in the area evaporates.

[0113] The predicted machine speed map generator 430 receives a prediction model 426 modeling a relationship between weed characteristics and machine speed (such as a prediction model generated by the weed characteristics to speed model generator 420) and generates a functional predicted machine speed map 438 of predicted desired machine speed at different locations in the field based on vegetation index values in the vegetation index map 332, or predicted weed intensity values in the predicted weed map 360, or vegetation values in the prior operation map 400 at those locations in the field and based on the prediction model 426.

[0114] The predicted operator command map generator 432 receives a prediction model 426 modeling a relationship between weed characteristics and operator command inputs detected by the operator input sensors 404 (such as a prediction model generated by the weed characteristics to command model generator 422) and generates a functional predicted operator command map 440 of predicted operator command inputs at different locations in the field based on vegetation index values from the vegetation index map 432, weed intensity or weed type values from the functional predicted weed map 360, or vegetation values from the prior operation map 400 and based on the prediction model 426.

[0115] The function prediction generator 212 outputs one or more of the function prediction maps 436, 437, 438, and 440. Each of the function prediction maps 436, 437, 438, and 440 can be provided to the control zone generator 213, the control system 214, or both. The control zone generator 213 generates control zones to provide a prediction control zone map 265 corresponding to each map 436, 437, 438, and 440 received by the control zone generator 213. Any or all of the function prediction maps 436, 437, 438, or 440 and the corresponding map 265 can be provided to the control system 214, which generates control signals to control one or more of the controllable subsystems 216 based on one or all of the function prediction maps 436, 437, 438, and 440 or the corresponding map 265 including the control zones. Any or all of the maps 436, 437, 438, or 440 or the corresponding map 265 can be presented to the operator 260 or another user.

[0116] Figure 6B is a block diagram showing some examples of real-time (live) sensors 208. Figure 6B Some or different combinations of the sensors shown in can have the sensors 336 and the processing system 338 at the same time. Figure 6B Some of the possible live sensors 208 shown in are shown and described with respect to the previous figure and are similarly numbered. Figure 6B It is shown that the live sensors 208 can include operator input sensors 980, machine sensors 982, harvested material property sensors 984, field and soil property sensors 985, environmental characteristic sensors 987, and they can include a wide variety of other sensors 226. The non-machine sensors 983 include the operator input sensor(s) 980, the harvested material property sensor(s) 984, the field and soil property sensor(s) 985, the environmental characteristic sensor(s) 987, and can also include other sensors 226. The operator input sensors 980 can be sensors that sense operator input through the operator interface mechanisms 218. Thus, the operator input sensors 980 can sense user movements of a joystick, a joystick, a steering wheel, a button, a dial, or a pedal. The operator input sensors 980 can also sense user interaction with other operator input structures, such as interaction with a touch-sensitive screen, with a microphone that utilizes voice recognition, or any of a variety of other operator input mechanisms.

[0117] Machine sensor 982 can sense various characteristics of the agricultural harvester 100. For example, as discussed above, machine sensor 982 may include machine speed sensor 146, separator loss sensor 148, clean grain camera 150, forward-view image capture mechanism 151, loss sensor 152, or geolocation sensor 204, examples of which are described above. Machine sensor 982 may also include machine setting sensor 991 that senses machine settings. (See above references) Figure 1Some examples of machine settings are described. A front equipment (e.g., header) position sensor 993 can sense a position of the header 102, reel 164, cutter 104, or other front end equipment relative to a frame of the agricultural harvester 100. For example, the sensor 993 can sense a height of the header 102 above the ground. The machine sensors 982 can also include a front equipment (e.g., header) orientation sensor 995. The sensor 995 can sense an orientation of the header 102 relative to the agricultural harvester 100 or relative to the ground. The machine sensors 982 can include a stability sensor 997. The stability sensor 997 senses a vibration or bounce motion (and amplitude) of the agricultural harvester 100. The machine sensors 982 can also include a residue setting sensor 999 configured to sense whether the agricultural harvester 100 is configured to chop residue, produce a windrow, or process residue in another manner. The machine sensors 982 can include a clean grain bin fan speed sensor 951 that senses a speed of the clean grain fan 120. The machine sensors 982 can include a concave gap sensor 953 that senses a gap between the rotor 112 and the concave 114 on the agricultural harvester 100. The machine sensors 982 can include a chaffer gap sensor 955 that senses a size of openings in the chaffer sieve 122. The machine sensors 982 can include a threshing rotor speed sensor 957 that senses a rotor speed of the rotor 112. The machine sensors 982 can include a rotor pressure sensor 959 that senses a pressure used to drive the rotor 112. The machine sensors 982 can include a screen gap sensor 961 that senses a size of openings in the screen 124. The machine sensors 982 can include a MOG moisture sensor 963 that senses a moisture level of MOG passing through the agricultural harvester 100. The machine sensors 982 can include a machine orientation sensor 965 that senses an orientation of the agricultural harvester 100. The machine sensors 982 can include a material feed rate sensor 967 that senses a feed rate of material as it travels through the feedhouse 106, clean grain elevator 130, or elsewhere in the agricultural harvester 100. The machine sensors 982 can include a biomass sensor 969 that senses biomass traveling through the feedhouse 106, the separator 116, or elsewhere in the agricultural harvester 100. The machine sensors 982 can include a fuel consumption sensor 971 that senses a fuel consumption rate of the agricultural harvester 100 over time. The machine sensors 982 can include a power utilization sensor 973 that senses power utilization in the agricultural harvester 100 (such as which subsystems are utilizing power), or a rate at which the subsystems are utilizing power, or a power distribution between subsystems in the agricultural harvester 100. The machine sensors 982 can include a tire pressure sensor 977 that senses an inflation pressure in the tires 144 of the agricultural harvester 100. The machine sensors 982 can include a variety of other machine performance or machine characteristic sensors (as shown by block 975).Machine performance and machine characteristic sensors 975 can sense machine performance or characteristics of the agricultural harvester 100.

[0118] The harvested material properties sensor 984 can sense characteristics of the severed crop material as the crop material is being processed by the agricultural harvester 100. Crop properties can include things such as crop type, crop moisture, grain quality (e.g., broken grain), MOG levels, grain constituents (such as starch and protein), MOG moisture, and other crop material properties. Other sensors can sense stalk “toughness,” adhesion of corn to the ear, and other characteristics that can be beneficially used to control processing for better grain capture, reduced grain damage, reduced power consumption, reduced grain loss, and the like.

[0119] Field and soil properties sensor 985 can sense characteristics of the field and soil. Field and soil properties can include soil moisture, soil compaction, presence and location of standing water, soil type, and other soil and field characteristics.

[0120] Environmental characteristics sensor 987 can sense one or more environmental characteristics. Environmental characteristics can include things such as wind direction and speed, precipitation, fog, dust levels or other obscuring matter, or other environmental characteristics.

[0121] In some examples, Figure 6B One or more of the sensors illustrated in FIG. 10 are processed to receive processed data 409 and are used as inputs to a model generator 210. The model generator 210 generates a model indicative of a relationship between the sensor data and one or more of a priori information maps or predicted information maps. The model is provided to a map generator 212, which generates a map mapping predicted sensor data values or related characteristics corresponding to the sensors from Figure 6B FIG. 10.

[0122] Figure 7A flowchart showing one example of the operation of the predictive model generator 210 and the prediction map generator 212 in generating one or more predictive models 426 and one or more functional prediction maps 436, 437, 438, and 440 is shown. At block 442, the predictive model generator 210 and the prediction map generator 212 receive an information map. The information map can be the vegetation index map 332, the predictive weed map 360, or a prior operation map 400 created using data obtained during a prior operation in the field. At block 444, the predictive model generator 210 receives sensor signals containing sensor data from the field sensors 208. The field sensors can be one or more of the biomass sensors 402 (which can be optical sensors 410 or rotor pressure sensors 412), the machine speed sensor 146, or another sensor 414. Block 446 indicates that the sensor signals received by the predictive model generator 210 include data indicative of the type of biomass. Block 448 indicates that the sensor signal data can be indicative of the speed of the agricultural harvester 100. Block 450 indicates that the sensor signals received by the prediction map generator 210 can be sensor signals sensed by the operator input sensors 404 having data indicative of the type of operator command input. The predictive model generator 210 can also receive other field sensor inputs as indicated by block 452.

[0123] At block 454, the processing system 406 processes the data contained in the one or more sensor signals received from the one or more field sensors 208 to obtain processed data 409 as shown. Figure 6A The data contained in the one or more sensor signals can be in a raw format that is processed to receive the processed data 409. For example, a temperature sensor signal includes resistance data that can be processed into temperature data. In other examples, the processing can include digitizing, encoding, formatting, scaling, filtering, or classifying the data. The processed data 409 can be indicative of one or more of biomass, machine speed, or operator input commands. The processed data 409 is provided to the predictive model generator 210.

[0124] Returning to Figure 7 At block 456, the predictive model generator 210 also receives a geographic location 334 from the geographic location sensor 204 as shown. Figure 6A The geographic location 334 can be associated with the geographic location from which the sensed variable sensed by the field sensors 208 is obtained. For example, the predictive model generator 210 can obtain the geographic location 334 from the geographic location sensor 204 and determine the precise geographic location from which the processed data 409 is derived based on machine delay, machine speed, and the like.

[0125] At block 458, the prediction model generator 210 generates one or more prediction models 426 that model relationships between the mapped values in the information maps and the characteristics represented in the processed data 409. For example, in some cases, the mapped values in the information maps can be weed characteristics, which can be one or more of the vegetation index values in the vegetation index map 332, the weed intensity values in the functional predicted weeds map 360, or the vegetation values of the prior operation map 400; and the prediction model generator 210 generates prediction models using the mapped values of the information maps and the characteristics sensed by the field sensors 208 (as represented in the processed data 490) or related characteristics, such as characteristics related to the characteristics sensed by the field sensors 208.

[0126] For example, at block 460, the prediction model generator 210 can generate prediction models 426 that model relationships between vegetation index, weed intensity, or vegetation values obtained from one or more information maps and biomass data obtained from the field sensors. In another example, at block 462, the prediction model generator 210 can generate prediction models 426 that model relationships between vegetation index, weed intensity, or vegetation values obtained from one or more information maps and speed of the agricultural harvester 100 obtained from the field sensors. In yet another example, at block 464, the weed characteristics to operator command model generator 422 generates prediction models 426 that model relationships between weed characteristics and operator command inputs.

[0127] The one or more prediction models 426 are provided to the prediction map generator 212. At block 466, the prediction map generator 212 generates one or more functional prediction maps. The functional prediction maps can be a functional predicted biomass map 436, a functional predicted moisture map 437, a functional predicted machine speed map 438, a functional predicted operator command map 440, or any combination of these maps. The functional predicted biomass map 436 predicts the biomass that will be encountered by the agricultural harvester 100 at different locations in the field. The functional predicted moisture map 437 predicts the material moisture that is expected to be encountered by the agricultural harvester 100 at different locations in the field. The functional predicted machine speed map 438 predicts the machine speed that is expected by the agricultural harvester 100 at different locations in the field, and the functional predicted operator command map 440 can predict the operator command inputs at different locations in the field. Further, one or more of the functional prediction maps 436, 437, 438, and 440 can be generated during the course of the agricultural operation. Thus, as the agricultural harvester 100 moves through the field, thereby performing the agricultural operation, one or more of the prediction maps 436, 437, 438, and 440 are generated as the agricultural operation is performed.

[0128] At block 468, the prediction map generator 212 outputs one or more functional prediction maps 436, 437, 438, and 440. At block 470, the prediction map generator 212 can configure the maps for presentation to the operator 260 or another user and possible interaction by the operator 260 or another user. At block 472, the prediction map generator 212 can configure the maps for use by the control system 214. At block 474, the prediction map generator 212 can provide one or more of the prediction maps 436, 437, 438, and 440 to the control zone generator 213 for generating control zones. At block 476, the prediction map generator 212 otherwise configures one or more of the prediction maps 436, 437, 438, and 440. In an example where one or more of the functional prediction maps 436, 437, 438, and 440 are provided to the control zone generator 213, the one or more functional prediction maps 436, 437, 438, and 440 and control zones included therein (represented by the corresponding map 265, as described above) can be presented to the operator 260 or another user, or also provided to the control system 214.

[0129] At block 478, the control system 214 then generates control signals to control the controllable subsystems based on one or more of the functional prediction maps 436, 437, 438, and 440 (or the functional prediction maps 436, 437, 438, and 440 with control zones) and inputs from the geographic location sensor 204. For example, when the functional prediction biomass map 436 or the functional prediction biomass map 436 with control zones is provided to the control system 214, in response, the feed rate controller 236 generates control signals to control the controllable subsystem 216 to control the feed rate of material through the agricultural harvester 100 based on the predicted biomass level. This is indicated by block 480.

[0130] Block 482 illustrates an example where the control system 214 receives the functional prediction machine speed map 438 or the functional prediction machine speed map 438 with control zones added. In response, the settings controller 232 controls the propulsion subsystem 250 (shown in Figure 2 as one of the controllable subsystems 216) to control the speed of the agricultural harvester 100 based on the predicted speed values in the functional prediction machine speed map 438 or the functional prediction machine speed map 438 with control zones.

[0131] Block 484 illustrates an example where the functional prediction operator command map 440 is provided to the control system 214. In response, the settings controller 232 generates control signals to control the controllable subsystems 216 to automatically generate command inputs based on the operator command values in the functional prediction operator command map 440 or the functional prediction operator command map 440 with control zones, or to recommend command inputs to the operator 260.

[0132] Figure 8 A block diagram illustrating one example of a control zone generator 213 is shown. The control zone generator 213 includes a work machine actuator (WMA) selector 486, a control zone generation system 488, and a dynamic regime generation system 490. The control zone generator 213 can also include other items 492. The control zone generation system 488 includes a control zone criteria identifier component 494, a control zone boundary definition component 496, a target setting identifier component 498, and other items 520. The dynamic regime generation system 490 includes a dynamic regime criteria identification component 522, a dynamic regime boundary definition component 524, a setting resolver identifier component 526, and other items 528. Before describing the overall operation of the control zone generator 213 in more detail, a brief description of some of the items in the control zone generator 213 and their respective operations will first be provided.

[0133] The agricultural harvester 100 or other work machine can have multiple different types of controllable actuators that perform different functions. The controllable actuators on the agricultural harvester 100 or other work machine are collectively referred to as work machine actuators (WMAs). Each WMA can be independently controlled based on values on a function prediction map, or the WMAs can be controlled in groups based on one or more values on the function prediction map. Thus, the control zone generator 213 can generate control zones corresponding to each individually controllable WMA, or to groups of WMAs that are controlled in coordination with each other.

[0134] The WMA selector 486 selects a WMA or group of WMAs for which a corresponding control zone is to be generated. The control zone generation system 488 then generates a control zone for the selected WMA or group of WMAs. Different criteria can be used in identifying the control zone for each WMA or group of WMAs. For example, for one WMA, WMA response time can be used as a criterion for defining boundaries of a control zone. In another example, wear characteristics (e.g., how much a particular actuator or mechanism wears out due to its movement) can be used as a criterion for identifying boundaries of a control zone. The control zone criteria identifier component 494 identifies the particular criteria that will be used to define a control zone for the selected WMA or group of WMAs. The control zone boundary definition component 496 processes values on the function prediction map in the analysis to define boundaries of a control zone on the function prediction map in the analysis based on the values on the function prediction map in the analysis and based on the control zone criteria for the selected WMA or group of WMAs.

[0135] The target setting identifier component 498 sets the value of the target setting that will be used to control the WMA or WMA group in the different control zones. For example, if the selected WMA is the propulsion system 250 and the function prediction map under analysis is the function prediction speed map 438, the target setting in each control zone can be a target speed setting based on the speed values contained in the function prediction speed map 238 within the identified control zone.

[0136] In some examples, where the agricultural harvester 100 is controlled based on the current or future location of the agricultural harvester 100, multiple target settings are possible for a WMA at a given location. In this case, the target settings can have different values and can compete with each other. Therefore, the target settings need to be resolved so that only a single target setting is used to control the WMA. For example, where the WMA is an actuator that is controlled in the propulsion system 250 in order to control the speed of the agricultural harvester 100, there can be multiple different competing sets of criteria that are considered by the control zone generation system 488 when identifying the control zones and the target settings for the selected WMA in the control zones. For example, different target settings for controlling the speed of the machine can be generated based on, for example, a detected or predicted feed rate value, a detected fuel efficiency value or a predicted fuel efficiency value, a detected or predicted grain loss value, or a combination of these values. However, at any given time, the agricultural harvester 100 cannot travel over the ground at multiple speeds simultaneously. Rather, at any given time, the agricultural harvester 100 travels at a single speed. Therefore, one of the competing target settings is selected to control the speed of the agricultural harvester 100.

[0137] Therefore, in some examples, the dynamic zone generation system 490 generates dynamic zones to resolve the multiple different competing target settings. The dynamic zone criteria identification component 522 identifies criteria for establishing a dynamic zone for the selected WMA or WMA group on the function prediction map under analysis. Some criteria that can be used to identify or define a dynamic zone include, for example, crop type or crop class based on the planting map, or another source of crop type or crop class, weed type, weed intensity, or crop status such as whether the crop is laid flat, partially laid flat, or standing. Just as each WMA or WMA group can have a corresponding control zone, different WMA or WMA groups can have a corresponding dynamic zone. The dynamic zone boundary definition component 524 identifies the boundaries of the dynamic zone on the function prediction map under analysis based on the dynamic zone criteria identified by the dynamic zone criteria identification component 522.

[0138] In some examples, the dynamic zones can overlap one another. For example, a crop class dynamic zone can overlap some or all of a crop status dynamic zone. In such examples, different dynamic zones can be assigned to a priority hierarchy such that, in the event of overlap of two or more dynamic zones, the dynamic zone assigned a higher position or importance in the priority hierarchy is prioritized over the dynamic zone with a lower position or importance in the priority hierarchy. The priority hierarchy of the dynamic zones can be set manually or can be set automatically using a rules-based system, a model-based system, or other system. As one example, in the event of overlap of a lodged crop dynamic zone and a crop class dynamic zone, the lodged crop dynamic zone can be assigned a greater importance in the priority hierarchy than the crop class dynamic zone such that the lodged crop dynamic zone is prioritized.

[0139] Further, for a given WMA or group of WMAs, each dynamic zone can have a unique settings resolver. The settings resolver identifier component 526 identifies a particular settings resolver for each dynamic zone identified on the functional prediction map under analysis and identifies the particular settings resolver for the selected WMA or group of WMAs.

[0140] Once the settings resolver for a particular dynamic zone is identified, the settings resolver can be used to resolve a competing target setting in which more than one target setting is identified based on the control zone. Different types of settings resolvers can have different forms. For example, the settings resolver identified for each dynamic zone can include a human selection resolver in which the competing target setting is presented to an operator or other user for resolution. In another example, the settings resolver can include a neural network or other artificial intelligence or machine learning system. In such a case, the settings resolver can resolve the competing target setting based on a predicted quality metric or historical quality metric corresponding to each of the different target settings. As an example, an increased vehicle speed setting can reduce the time to harvest a field and reduce the corresponding time-based labor and equipment costs, but can increase grain loss. A decreased vehicle speed setting can increase the time to harvest a field and increase the corresponding time-based labor and equipment costs, but can reduce grain loss. When grain loss or harvesting time is selected as the quality metric, the predicted value or historical value of the selected quality metric can be used to resolve the speed setting given two competing vehicle speed setting values. In certain cases, the settings resolver can be a set of threshold rules that can be used in place of or in addition to the dynamic zones. An example of a threshold rule can be expressed as follows:

[0141] If the predicted biomass value within 20 feet of header 20 of agricultural harvester 100 is greater than x kilograms (where x is a selected or predetermined value), then the target setting value based on the feed rate rather than other competing target setting values is used, otherwise the target setting value based on the grain loss rather than other competing target setting values is used.

[0142] The setting resolver can be a logical component that executes logical rules in identifying the target setting. For example, the setting resolver can resolve the target setting while attempting to minimize the harvesting time or minimize the total harvesting cost or maximize the harvested grain, or other variables computed as a function of different candidate target settings. The harvesting time can be minimized when the amount of harvesting completed is reduced to or below a selected threshold. The total harvesting cost can be minimized when the total harvesting cost is reduced to or below a selected threshold. The harvested grain can be maximized when the amount of harvested grain is increased to or above a selected threshold.

[0143] Figure 9 is a flowchart illustrating one example of the operation of control zone generator 213 in generating control zones and dynamic zones for control zone generator 213 receiving a graph for zone processing, e.g., a graph in analysis.

[0144] At block 530, control zone generator 213 receives a graph in analysis for processing. In one example, as shown at block 532, the graph in analysis is a functional prediction graph. For example, the graph in analysis can be one of functional prediction graphs 436, 437, 438, or 440. Block 534 indicates that the graph in analysis can also be other graphs.

[0145] At block 536, the WMA selector 486 selects a WMA or group of WMAs for which a control zone is to be generated on the map in the analysis. At block 538, the control zone criteria identification component 494 obtains control zone definition criteria for the selected WMA or group of WMAs. Block 540 indicates such an example in which the control zone criteria are or include wear characteristics of the selected WMA or group of WMAs. Block 542 indicates such an example in which the control zone definition criteria are or include magnitudes and variations of input source data, such as magnitudes and variations of values on the map in the analysis or magnitudes and variations of inputs from various field sensors 208. Block 544 indicates such an example in which the control zone definition criteria are or include physical machine characteristics, such as physical dimensions of the machine, speeds of different subsystem operations, or other physical machine characteristics. Block 546 indicates such an example in which the control zone definition criteria are or include responsiveness of the selected WMA or group of WMAs in reaching set values of new commands. Block 548 indicates such an example in which the control zone definition criteria are or include machine performance metrics. Block 550 indicates such an example in which the control zone definition criteria are or include operator preferences. Block 552 indicates such an example in which the control zone definition criteria are also or include other items. Block 549 indicates such an example in which the control zone definition criteria are time-based, meaning that the agricultural harvester 100 will not cross the boundary of a control zone until a selected amount of time has elapsed since the agricultural harvester 100 entered the particular control zone. In some cases, the selected amount of time can be a minimum amount of time. Thus, in some cases, the control zone definition criteria can prevent the agricultural harvester 100 from crossing the boundary of a control zone until at least the selected amount of time has elapsed. Block 551 indicates such an example in which the control zone definition criteria are based on a selected size value. For example, control zone definition criteria based on a selected size value can exclude definition of control zones that are smaller than the selected size. In some cases, the selected size can be a minimum size.

[0146] At block 554, the dynamic zone criteria identification component 522 obtains dynamic zone definition criteria for the selected WMA or group of WMAs. Block 556 indicates such an example in which the dynamic zone definition criteria are based on manual input from the operator 260 or another user. Block 558 shows such an example in which the dynamic zone definition criteria are based on crop type or crop class. Block 560 shows such an example in which the dynamic zone definition criteria are based on weed type or weed intensity or both. Block 562 shows such an example in which the dynamic zone definition criteria are based on or include crop status. Block 564 indicates such an example in which the dynamic zone definition criteria are also or include other criteria.

[0147] At block 566, control zone boundary definition component 496 generates boundaries of control zones on the map under analysis based on the control zone criteria. Dynamic zone boundary definition component 524 generates boundaries of dynamic zones on the map under analysis based on the dynamic zone criteria. Block 568 indicates examples in which zone boundaries are identified for both control zones and dynamic zones. Block 570 shows that target setting identifier component 498 identifies target settings for each of the control zones. Control zones and dynamic zones can also be generated in other ways, and this is indicated by block 572.

[0148] At block 574, setting resolver identifier component 526 identifies a setting resolver for the selected WMA in each dynamic zone defined by dynamic zone boundary definition component 524. As discussed above, the dynamic zone resolver can be a human resolver 576, an artificial intelligence or machine learning system resolver 578, a resolver based on a predicted or historical quality of each competing target setting 580, a rule-based resolver 582, a performance criteria-based resolver 584, or other resolver 586.

[0149] At block 588, WMA selector 486 determines whether there are more WMAs or groups of WMAs to process. If there are additional WMAs or groups of WMAs to process, processing returns to block 436 in which the next WMA or group of WMAs for which control zones and dynamic zones are to be defined is selected. When there are no additional WMAs or groups of WMAs for which control zones or dynamic zones are to be generated left, processing moves to block 590 in which control zone generator 213 outputs a map for each of the WMA or groups of WMAs with control zones, target settings, dynamic zones, and setting resolvers. As discussed above, the output map can be presented to operator 260 or another user; the output map can be provided to control system 214; or the output map can be output in other ways.

[0150] Figure 10 One example is shown in which control system 214 controls operation of agricultural harvester 100 based on the map output by control zone generator 213. Thus, at block 592, control system 214 receives the map of the work site. In some cases, the map can be a functional prediction map that can include control zones and dynamic zones (as shown in block 594). In some cases, the received map can be a functional prediction map that excludes control zones and dynamic zones. Block 596 indicates examples in which the received map of the work site can be an information map with control zones and dynamic zones identified thereon. Block 598 indicates examples in which the received map can include multiple different maps or multiple different map layers. Block 610 indicates examples in which the received map can also take other forms.

[0151] At block 612, the control system 214 receives a sensor signal from the geo-location sensor 204. The sensor signal from the geo-location sensor 204 can include data indicative of a geographic location 614 of the agricultural harvester 100, a speed 616 of the agricultural harvester 100, a heading 618 of the agricultural harvester 100, or other information 620. At block 622, the zone controller 247 selects a dynamic zone, and at block 624, the zone controller 247 selects a control zone on the map based on the geo-location sensor signal. At block 626, the zone controller 247 selects a WMA or group of WMAs to be controlled. At block 628, the zone controller 247 obtains one or more target settings for the selected WMA or group of WMAs. The target settings obtained for the selected WMA or group of WMAs can come from a variety of different sources. For example, block 630 illustrates an example in which one or more of the target settings for the selected WMA or group of WMAs are based on input from the control zone on the map from the job site. Block 632 illustrates an example in which one or more of the target settings are obtained from manual input by the operator 260 or another user. Block 634 illustrates an example in which the target settings are obtained from the field sensors 208. Block 636 illustrates an example in which one or more of the target settings are obtained from one or more sensors on other machines simultaneously operating in the same field as the agricultural harvester 100 or from one or more sensors on machines that have operated in the same field in the past. Block 638 illustrates an example in which the target settings are also obtained from other sources.

[0152] At block 640, the zone controller 247 accesses the setting resolver for the selected dynamic zone and controls the setting resolver to resolve the competing target settings into resolved target settings. As discussed above, in some cases, the setting resolver can be a human resolver, in which case the zone controller 247 controls the operator interface mechanism 218 to present the competing target settings to the operator 260 or another user for resolution. In some cases, the setting resolver can be a neural network or other artificial intelligence or machine learning system, and the zone controller 247 submits the competing target settings to the neural network, artificial intelligence, or machine learning system for selection. In certain cases, the setting resolver can be based on predicted quality metrics or historical quality metrics, based on threshold rules, or based on a logic component. In any of these latter examples, the zone controller 247 executes the setting resolver to obtain the resolved target settings based on predicted quality metrics or historical quality metrics, based on threshold rules, or in the case of using a logic component.

[0153] At block 642, with the resolved target settings identified by the zone controller 247, the zone controller 247 provides the resolved target settings to other controllers in the control system 214, which generate control signals based on the resolved target settings and apply the control signals to the selected WMA or WMA group. For example, with the selected WMA being a machine or header actuator 248, the zone controller 247 provides the resolved target settings to the settings controller 232 or the header / actual controller 238 or both to generate control signals based on the resolved target settings, and those generated control signals are applied to the machine or header actuator 248. At block 644, if additional WMAs or additional WMA groups are to be controlled at the current geographic location of the agricultural harvester 100 (as detected at block 612), the process returns to block 626, where the next WMA or WMA group is selected. The process represented by blocks 626 through 644 continues until all WMAs or WMA groups to be controlled at the current geographic location of the agricultural harvester 100 have been addressed. If no additional WMAs or WMA groups remain to be controlled at the current geographic location of the agricultural harvester 100, the process proceeds to block 646, where the zone controller 247 determines whether additional control zones to be considered exist in the selected dynamic zone. If additional control zones to be considered exist, the process returns to block 624, where the next control zone is selected. If no additional control zones need to be considered, the process proceeds to block 648, where a determination is made as to whether additional dynamic zones need to be considered. The zone controller 247 determines whether additional dynamic zones need to be considered. If additional dynamic zones need to be considered, the process returns to block 622, where the next dynamic zone is selected.

[0154] At block 650, the zone controller 247 determines whether the operation being performed by the agricultural harvester 100 is complete. If not, the zone controller 247 determines whether the control zone criteria have been met to continue processing, as shown at block 652. For example, as mentioned above, the control zone definition criteria can include criteria defining when the agricultural harvester 100 can cross the control zone boundary. For example, whether the agricultural harvester 100 can cross the control zone boundary can be defined by a selected time period, meaning that the agricultural harvester 100 is prevented from crossing the zone boundary until the selected amount of time has elapsed. In this case, at block 652, the zone controller 247 determines whether the selected time period has elapsed. Additionally, the zone controller 247 can continuously perform the processing. Thus, the zone controller 247 does not wait for any particular time period before continuing to determine whether the operation of the agricultural harvester 100 is complete. At block 652, the zone controller 247 determines that it is time to continue processing, and then processing continues at block 612, where the zone controller 247 again receives input from the geo-location sensor 204. It should also be understood that the zone controller 247 can use a multiple-input, multiple-output controller to simultaneously control the WMAs and WMA groups, rather than sequentially controlling the WMAs and WMA groups.

[0155] Figure 11 is a block diagram illustrating one example of an operator interface controller 231. In the illustrated example, the operator interface controller 231 includes an operator input command processing system 654, other controller interaction system 656, a speech processing system 658, and an action signal generator 660. The operator input command processing system 654 includes a speech handling system 662, a touch gesture processing system 664, and other items 666. The other controller interaction system 656 includes a controller input processing system 668 and a controller output generator 670. The speech processing system 658 includes a trigger detector 672, a recognition component 674, a synthesis component 676, a natural language understanding system 678, a dialog management system 680, and other items 682. The action signal generator 660 includes a visual control signal generator 684, an audio control signal generator 686, a haptic control signal generator 688, and other items 690. A brief description of some of the items in the operator interface controller 231 and their associated operations is provided before describing the operation of the example operator interface controller 231 in handling various operator interface actions. Figure 11

[0156] ​The operator input command processing system 654 detects operator inputs on the operator interface mechanisms 218 and processes these command inputs. The speech handling system 662 detects speech inputs and processes interactions with the speech processing system 658 to process speech command inputs. The touch gesture processing system 664 detects touch gestures on touch sensitive elements in the operator interface mechanisms 218 and processes these command inputs.

[0157] Other controller interaction system 656 processes interactions with other controllers in the control system 214. The controller input processing system 668 detects and processes inputs from other controllers in the control system 214 and the controller output generator 670 generates outputs and provides these outputs to other controllers in the control system 214. The speech processing system 658 recognizes speech inputs, determines the meaning of these inputs, and provides outputs indicating the meaning of the verbal inputs. For example, the speech processing system 658 can recognize a speech input from the operator 260 as a set change command where the operator 260 is commanding the control system 214 to change a setting of a controllable subsystem 216. In such an example, the speech processing system 658 recognizes the content of the verbal command, identifies the meaning of the command as a set change command, and provides the meaning of the input back to the speech handling system 662. The speech handling system 662 in turn interacts with the controller output generator 670 to provide command outputs to the appropriate controller in the control system 214 to complete the verbal set change command.

[0158] The voice processing system 658 can be invoked in a variety of different ways. For example, in one example, the voice handling system 662 continuously provides input from a microphone (as one of the operator interface mechanisms 218) to the voice processing system 658. The microphone detects voice from the operator 260, and the voice handling system 662 provides the detected voice to the voice processing system 658. The trigger detector 672 detects a trigger that indicates that the voice processing system 658 is invoked. In some cases, when the voice processing system 658 receives continuous voice input from the voice handling system 662, the speech recognition component 674 performs continuous speech recognition on all speech spoken by the operator 260. In some cases, the voice processing system 658 is configured to be invoked using a wake-up word. That is, in some cases, the operation of the voice processing system 658 can be initiated based on recognizing a selected spoken word, referred to as a wake-up word. In such examples, where the recognition component 674 recognizes the wake-up word, the recognition component 674 provides an indication to the trigger detector 672 that the wake-up word has been recognized. The trigger detector 672 detects that the voice processing system 658 has been invoked or triggered by the wake-up word. In another example, the voice processing system 658 can be invoked by the operator 260 actuating an actuator on a user interface mechanism, such as by touching an actuator on a touch-sensitive display screen, by pressing a button, or by providing another trigger input. In such examples, when the trigger input via the user interface mechanism is detected, the trigger detector 672 can detect that the voice processing system 658 has been invoked. The trigger detector 672 can also detect that the voice processing system 658 has been invoked in other ways.

[0159] Once the voice processing system 658 is invoked, voice input from the operator 260 is provided to the speech recognition component 674. The speech recognition component 674 recognizes linguistic elements in the voice input, such as words, phrases, or other linguistic units. The natural language understanding system 678 identifies a meaning of the recognized voice. The meaning can be any of a natural language output, a command output identifying a command reflected in the recognized voice, a value output identifying a value in the recognized voice, or a variety of other outputs reflecting an understanding of the recognized voice. For example, more generally, the natural language understanding system 678 and the voice processing system 568 can understand a meaning of voice recognized in the context of the agricultural harvester 100.

[0160] In some examples, the speech processing system 658 can also generate output that is presented to the user through the user experience navigation operator 260 based on the speech input. For example, the dialog management system 680 can generate and manage a dialog with the user in order to identify what the user wants to do. The dialog box can disambiguate user commands, identify one or more particular values needed to perform the user command; or obtain other information from the user or provide other information to the user or both. The synthesis component 676 can generate speech synthesis that can be presented to the user through an audio operator interface mechanism such as a speaker. Thus, the dialog managed by the dialog management system 680 can be exclusively a spoken dialog or a combination of a visual dialog and a spoken dialog.

[0161] The action signal generator 660 generates action signals to control the operator interface mechanisms 218 based on output from one or more of the operator input command processing system 654, the other controller interaction system 656, and the speech processing system 658. The visual control signal generator 684 generates control signals to control visual items in the operator interface mechanisms 218. The visual items can be lights, display screens, warning indicators, or other visual items. The audio control signal generator 686 generates output to control audio elements of the operator interface mechanisms 218. The audio elements include speakers, audible alert mechanisms, horns, or other audible elements. The haptic control signal generator 688 generates control signals that are output to control haptic elements of the operator interface mechanisms 218. The haptic elements include vibratory elements that can be used to vibrate, for example, the operator's seat, steering wheel, pedals, or joystick used by the operator. The haptic elements can include tactile feedback or force feedback elements that provide tactile or force feedback to the operator through the operator interface mechanisms. The haptic elements can also include a wide variety of other haptic elements.

[0162] Figure 12 is a flowchart showing one example of the operation of the operator interface controller 231 in generating an operator interface display on the operator interface mechanisms 218, which can include a touch-sensitive display screen. Figure 12 One example of how the operator interface controller 231 can detect and process operator interaction with the touch-sensitive display screen is also shown.

[0163] At block 692, the operator interface controller 231 receives a map. Block 694 indicates an example in which the map is a function prediction map, while block 696 indicates an example in which the map is another type of map. At block 698, the operator interface controller 231 receives input from the geo-location sensor 204 identifying a geo-location of the agricultural harvester 100. As shown in block 700, the input from the geo-location sensor 204 can include a heading and a position of the agricultural harvester 100. Block 702 indicates an example in which the input from the geo-location sensor 204 includes a speed of the agricultural harvester 100, while block 704 indicates an example in which the input from the geo-location sensor 204 includes other items.

[0164] At block 706, the visual control signal generator 684 in the operator interface controller 231 controls the touch-sensitive display screen in the operator interface mechanism 218 to generate a display showing all or a portion of the field represented by the received map. Block 708 indicates that the displayed field can include a current position marker showing a current position of the agricultural harvester 100 relative to the field. Block 710 indicates an example in which the displayed field includes a next work unit marker identifying a next work unit (or area on the field) in which the agricultural harvester 100 is to operate. Block 712 indicates an example in which the displayed field includes a coming area display portion showing areas that have not yet been processed by the agricultural harvester 100, while block 714 indicates an example in which the displayed field includes a previously visited display portion representing areas of the field that have already been processed by the agricultural harvester 100. Block 716 indicates an example in which the displayed field displays various characteristics of the field having geo-referenced locations on the map. For example, if the received map is a weed map, the displayed field can show different weed types present in the field geo-referenced within the displayed field. The mapped characteristics can be shown in previously visited areas (as shown in block 714), coming areas (as shown in block 712), and next work units (as shown in block 710). Block 718 indicates an example in which the displayed field includes other items.

[0165] Figure 13 is a diagram showing one example of a user interface display 720 that can be generated on a touch-sensitive display screen. In other implementations, the user interface display 720 can be generated on other types of displays. The touch-sensitive display screen can be mounted in an operator's compartment of the agricultural harvester 100 or on a mobile device or elsewhere. Before continuing with the description of the flowchart shown in Figure 12 the user interface display 720 will be described.

[0166] In Figure 13In the example shown in FIG. 7, the user interface display 720 shows that the touch- sensitive display includes display features for operating the microphone 722 and the speaker 724. Thus, the touch-sensitive display can be communicably coupled to the microphone 722 and the speaker 724. The box 726 indicates that the touch-sensitive display can include a variety of user interface control actuators, such as buttons, keypads, soft keypads, links, icons, switches, etc. The operator 260 can actuate the user interface control actuators to perform various functions.

[0167] In Figure 13 In the example shown in FIG. 7, the user interface display 720 includes a field display portion 728 that displays at least a portion of the field in which the agricultural harvester 100 is operating. The field display portion 728 is shown with a current position marker 708 that corresponds to the current position of the agricultural harvester 100 in the portion of the field shown in the field display portion 728. In one example, the operator can control the touch-sensitive display to zoom in on a portion of the field display portion 728 or to pan or scroll the field display portion 728 to display different portions of the field. A next work unit 730 is shown as a field region directly in front of the current position marker 708 of the agricultural harvester 100. The current position marker 708 can also be configured to identify a direction of travel of the agricultural harvester 100, a speed of travel of the agricultural harvester 100, or both. In Figure 13 In the example shown in FIG. 7, the shape of the current position marker 708 provides an indication of the orientation of the agricultural harvester 100 within the field, which can be used as an indication of the direction of travel of the agricultural harvester 100.

[0168] The size of the next work unit 730 labeled on the field display portion 728 can vary based on a variety of different criteria. For example, the size of the next work unit 730 can vary based on the speed of travel of the agricultural harvester 100. Thus, when the agricultural harvester 100 is traveling faster, the area of the next work unit 730 can be larger than if the agricultural harvester 100 is traveling slower. The field display portion 728 is also shown displaying a previously visited region 714 and an upcoming region 712. The previously visited region 714 represents an area that has already been harvested, while the upcoming region 712 represents an area that still needs to be harvested. The field display portion 728 is also shown displaying different characteristics of the field. In Figure 13 In the example shown in FIG. 7, the graph being displayed is a weed graph. Thus, a plurality of different weed markers are displayed on the field display portion 728. A set of weed characteristic display markers 732 is shown in the previously visited region 714. A set of weed characteristic display markers 734 is also shown in the upcoming region 712, and a set of weed characteristic display markers 736 is shown in the next work unit 730. Figure 13The weed characteristic display markers 732, 734, and 736 are shown to be composed of different symbols. Each of the symbols represents a type of weed. In the example shown in FIG. 3, the @ symbol represents amaranth; the * symbol represents foxtail; and the # symbol represents horseweed. Thus, the field display portion 728 shows different types of weeds located at different areas within the field. As previously mentioned, the display marker 732 can be composed of different symbols, and as described below, the symbols can be any display feature, such as different colors, shapes, patterns, intensities, text, icons, or other display features.

[0169] In Figure 13 the example, the user interface display 720 also has a control display portion 738. The control display portion 738 allows the operator to view information and interact with the user interface display 720 in various ways.

[0170] The actuators and display markers in the portion 738 can be displayed as, for example, separate items, a fixed list, a scrollable list, a drop-down menu, or a drop-down list. In Figure 13 the example shown, the display portion 738 shows information for three different types of weeds corresponding to the three symbols mentioned above. The display portion 738 also includes a set of touch-sensitive actuators that the operator 260 can interact with by touching. For example, the operator 260 can touch a touch-sensitive actuator with a finger to activate the corresponding touch-sensitive actuator.

[0171] The flag column 739 shows flags that have been set automatically or manually. The flag actuators 740 allow the operator 260 to mark a location and then add information indicating the type of weed found at that location. For example, when the operator 260 actuates the flag actuator 740 by touching the flag actuator 740, the touch gesture processing system 664 in the operator interface controller 231 identifies the location as one in which amaranth is present. When the operator 260 touches the button 742, the touch gesture processing system 664 identifies the location as one in which foxtail is present. When the operator 260 touches the button 744, the touch gesture processing system 664 identifies the location as one in which horseweed is present. The touch gesture processing system 664 also controls the visual control signal generator 684 to add the symbol corresponding to the identified type of weed on the field display portion 728 at the location identified by the button 740, 742, or 744 before or after or during actuation by the operator.

[0172] Column 746 displays symbols corresponding to each weed type tracked on field display section 728. Column 748 shows identifiers (which can be text identifiers or other identifiers) that identify the weed type. Without limitation, the weed symbols in column 746 and the identifiers in column 748 can include any display features, such as different colors, shapes, patterns, intensities, text, icons, or other display features. Column 750 displays weed characteristic values. Figure 12 In the example shown, the weed characteristic value is a value representing weed density. The value displayed in column 750 can be a predicted value or a value measured by field sensor 208. The value in column 750 can include any of the weed properties contained within the weed intensity range, as well as any of the weed type value and other values. In one example, operator 260 can select a specific portion of field display section 728 for which the values ​​in column 750 will be displayed. Thus, the values ​​in column 750 can correspond to the values ​​in display sections 712, 714, or 730. Column 752 displays an action threshold. The action threshold in column 752 can be a threshold corresponding to the measured value in column 750. If the measured value in column 750 meets the corresponding action threshold in column 752, the control system 214 takes the action identified in column 754. In some cases, the measured value can satisfy the corresponding action threshold by meeting or exceeding it. In one example, operator 260 can select a threshold, for example, to change the threshold by touching the threshold in column 752. Once selected, operator 260 can change the threshold. The threshold in column 752 can be configured to perform a specified action when the measured value 750 exceeds, is equal to, or is less than the threshold.

[0173] Similarly, operator 260 can touch the action identifier in column 754 to change the action to be taken. Multiple actions can be taken when a threshold is met. For example, at the bottom of column 754, deceleration action 756 and fan increase action 758 are identified as actions to be taken if the measured value in column 750 meets the threshold in column 752.

[0174] The actions that can be set in column 754 can be any of a variety of different types of actions. For example, the actions can include a prohibit action that, when executed, prevents the agricultural harvester 100 from further harvesting in the area. The actions can include a mitigation activation that, when executed, performs a mitigation action, such as activating a weed seed crusher or a weed seed bagger. The actions can include a speed change action that, when executed, changes the speed of travel of the agricultural harvester 100 through the field. The actions can include a setting change action to change a setting of an internal actuator or another WMA or group of WMAs, or to implement a change in a setting of the header. These are just examples, and a wide variety of other actions are contemplated herein.

[0175] The display indicia shown on the user interface display 720 can be visually controlled. Visually controlling the interface display 720 can be performed to capture the attention of the operator 260. For example, the display indicia can be controlled to modify the intensity, color, or pattern of the displayed display indicia. Additionally, the display indicia can be controlled to flash. As an example, the described changes to the visual appearance of the display indicia are provided. Thus, other aspects of the visual appearance of the display indicia can be changed. Thus, the display indicia can be modified in a desired manner in various situations, such as to capture the attention of the operator 260.

[0176] The various functions that can be accomplished by the operator 260 using the user interface display 720 can also be accomplished automatically, such as by other controllers in the control system 214. For example, when a different type of weed is identified by the field sensors 208, the operator interface controller 231 can automatically add a marker at the current location of the agricultural harvester 100, which corresponds to the location of the encountered weed type, and generate a display in the marker column, a corresponding symbol in the symbol column, and a designator in the designator column 748. The operator interface controller 231 can also generate the measured values in column 750 and the threshold values in column 752 when a different weed type is identified. The operator interface controller 231 or another controller can also automatically identify the actions added to column 754.

[0177] Now returning to Figure 12The flowchart of Figure 7 continues the description of the operation of the operator interface controller 231. At block 760, the operator interface controller 231 detects an input that sets a flag and controls the touch-sensitive user interface display 720 to display the flag on the field display portion 728. The detected input can be an operator input (as indicated at 762) or an input from another controller (as indicated at 764). At block 766, the operator interface controller 231 detects a field sensor input that indicates a measured characteristic of the field from one of the field sensors 208. At block 768, the visual control signal generator 684 generates a control signal to control the user interface display 720 to display actuators for modifying the user interface display 720 and for modifying machine control. For example, block 770 indicates that one or more of the actuators for setting or modifying the values in the columns 739, 746, and 748 can be displayed. Thus, the user can set the flags and modify the characteristics of the flags. For example, the user can modify the weed type and the weed designator that correspond to the flag. Block 772 indicates that the action threshold value in column 752 is displayed. Block 776 indicates that the action in column 754 is displayed, and block 778 indicates that the measured field data in column 750 is displayed. Block 780 indicates that a variety of other information and actuators can also be displayed on the user interface display 720.

[0178] At block 782, the operator input command processing system 654 detects and processes operator inputs that correspond to interactions with the user interface display 720 performed by the operator 260. In the case where the user interface mechanism on which the user interface display 720 is displayed is a touch-sensitive display screen, the interaction inputs made by the operator 260 with the touch-sensitive display screen can be touch gestures 784. In some cases, the operator interaction inputs can be inputs using a click device 786 or other operator interaction inputs 788.

[0179] At block 790, the operator interface controller 231 receives a signal indicating an alarm condition. For example, block 792 indicates that a signal can be received by the controller input processing system 668 indicating that the detected value in column 750 satisfies a threshold condition present in column 752. As explained previously, the threshold condition can include a value below, at, or above a threshold value. Block 794 shows that the action signal generator 660 can respond to receiving the alarm condition by generating a visual alarm using the visual control signal generator 684, generating an audio alarm using the audio control signal generator 686, generating a haptic alarm using the haptic control signal generator 688, or warning the operator 260 using any combination of these. Similarly, as shown in block 796, the controller output generator 670 can generate outputs to other controllers in the control system 214 so that these controllers perform the corresponding actions identified in column 754. Block 798 shows that the operator interface controller 231 can also detect and process alarm conditions in other ways.

[0180] Block 900 shows that the voice handling system 662 can detect and process inputs that invoke the voice processing system 658. Block 902 shows that performing voice processing can include using the dialog management system 680 to have a conversation with the operator 260. Block 904 shows that the voice processing can include providing signals to the controller output generator 670 so that control operations are automatically performed based on the voice input.

[0181] Table 1 below shows an example of a conversation between the operator interface controller 231 and the operator 260. In Table 1, the operator 260 invokes the voice processing system 658 using a trigger word or wake-up word that is detected by the trigger detector 672. In the example shown in Table 1, the wake-up word is “Johnny”.

[0182] Table 1

[0183] Operator: “Johnny, tell me about the current weeds”

[0184] Operator interface controller: “Sida herb 65%, threshold 10%. Crabgrass 15%, threshold 15%. Threeleaf sida 20%, threshold 25%”.

[0185] Operator: “Johnny, what should I do because of the weeds?”

[0186] Operator interface controller: “Sida herb is too tall. Stop harvesting in this area and reduce weed seed later”.

[0187] Table 2 shows an example in which the speech synthesis component 676 provides output to the audio control signal generator 686 to provide auditory updates intermittently or periodically. The interval between updates can be time-based (such as every five minutes), or coverage or distance-based (such as every five acres), or anomaly-based (such as when a measured value is greater than a threshold).

[0188] Table 2

[0189] Operator interface controller: "In the last 10 minutes, the harvest contains 90% crop, 5% amaranth, 4% threeleaf sowthistle, and 1% other."

[0190] Operator interface controller: "The next 1 acre of land contains 92% corn, 5% amaranth, and 3% other."

[0191] Operator interface controller: "Warning: amaranth is now 12%. Stop harvesting this area."

[0192] Operator interface controller: "Caution: threeleaf sowthistle is now 27%. Reduce grain cleaning fan speed by 200 rpm."

[0193] The example shown in Table 3 shows that some actuators or user input mechanisms on the touch-sensitive display 720 can be supplemented with voice dialog. The example in Table 3 shows that the action signal generator 660 can generate action signals to automatically flag amaranth weed patches in the field being harvested.

[0194] Table 3

[0195] Human: "Johnny, flag amaranth weed patches."

[0196] Operator interface controller: "Amaranth weed patches flagged."

[0197] The example shown in Table 4 shows that the action signal generator 660 can have a dialog with the operator 260 to start and stop flagging of weed patches.

[0198] Table 4

[0199] Human: "Johnny, start flagging amaranth weed patches."

[0200] Operator interface controller: "Flag amaranth weed patches."

[0201] Human: "Johnny, stop flagging amaranth weed patches."

[0202] Operator interface controller: "Amaranth weed patch flagging stopped."

[0203] The example shown in Table 5 shows that the action signal generator 160 can generate the signal marking the weed patch in a different manner than shown in Tables 3 and 4.

[0204] Table 5

[0205] Human: "Johnny, mark the next 100 feet as a patch of alligator weed."

[0206] Operator interface controller: "The next 100 feet is marked as a patch of alligator weed."

[0207] Returning again Figure 14 to FIG. 9, block 906 shows that the operator interface controller 231 can also detect and handle situations for outputting messages or other information in other ways. For example, the other controller interaction system 656 can detect input from other controllers indicating that an alert or output message should be presented to the operator 260. Block 908 shows that the output can be an audio message. Block 910 shows that the output can be a visual message, and block 912 shows that the output can be a haptic message. Until the operator interface controller 231 determines that the current harvesting operation is complete, as shown in block 914, the process returns to block 698, where the geographic position of the harvester 100 is updated, and the process continues as described above to update the user interface display 720.

[0208] Once the operation is complete, any desired values that were displayed or have been displayed on the user interface display 720 can be saved. These values can also be used in machine learning to improve different parts of the prediction model generator 210, the prediction map generator 212, the control zone generator 213, the control algorithm, or other items. The saved desired values are indicated by block 916. These values can be saved locally on the agricultural harvester 100, or the values can be saved at a remote server location or sent to another remote system.

[0209] As can be seen, the a priori information index map is obtained by the agricultural harvester and shows values of weed characteristics at different geographic locations of the field being harvested. On-board sensors on the harvester sense characteristics having values indicative of agricultural characteristics as the agricultural harvester moves through the field. The prediction map generator generates a prediction map that predicts control values for different locations in the field based on the values of the weed characteristics in the information map and the agricultural characteristics sensed by the on-board sensors. The control system controls the controllable subsystems based on the control values in the prediction map.

[0210] A control value is a value upon which an action can be based. As described herein, a control value can include any value (or a characteristic indicated by or derived from that value) that can be used to control the agricultural harvester 100. A control value can be any value indicative of an agricultural characteristic. A control value can be a predicted value, a measured value, or a detected value. A control value can include any of the characteristics indicated by or derived from a value provided by a graph, such as any one of the graphs described herein, for example, a control value can be a value provided by an information graph, a value provided by a prior information graph, or a value provided by a prediction graph, such as a functional prediction graph. A control value can also include any of the characteristics indicated by or derived from a value detected by any one of the sensors described herein. In other examples, a control value can be provided by an operator of the agricultural machine, such as a command entered by an operator of the agricultural machine.

[0211] The current discussion has mentioned processors and servers. In some examples, the processors and servers include computer processors with associated memory and timing circuitry, not separately shown. The processors and servers are a functional part of the systems or devices of which they are a part, and are activated by, and facilitate the functionality of, the other components or items in those systems.

[0212] Also, a number of user interface displays have been discussed. The displays can take a variety of different forms and can have a variety of different user-actuatable operator interface structures disposed thereon. For example, the user-actuatable operator interface structures can be text boxes, check boxes, icons, links, drop-down menus, search boxes, etc. The user-actuatable operator interface structures can also be actuated in a variety of different ways. For example, the user-actuatable operator interface structures can be actuated using an operator interface mechanism such as a pointing device, (such as a trackball or mouse, a hardware button, a switch, a joystick or keyboard, a thumb switch or thumb pad, etc.), a virtual keyboard or other virtual actuator. Further, where the screen on which the user-actuatable operator interface structures are displayed is a touch-sensitive screen, the user-actuatable operator interface structures can be actuated using touch gestures. Also, the user-actuatable operator interface structures can be actuated using voice commands using voice recognition functionality. Voice recognition can be implemented using a voice detection device such as a microphone and software for recognizing detected voice and executing commands based on received voice.

[0213] A number of data stores have also been discussed. It should be noted that the data stores can each be divided into a plurality of data stores. In some examples, one or more of the data stores can be local to the system accessing the data store, all of the data stores can be located remotely from the system utilizing the data stores, or one or more data stores can be local while others are remote. All of these configurations are contemplated by the present disclosure.

[0214] Furthermore, the attached figures illustrate a number of blocks that include functionality that is attributed to each block. It should be noted that the functionality attributed to multiple different blocks can be performed by fewer blocks, and that more functionality can be performed by a single block than the functionality attributed to a particular block. In different examples, some functionality can be added, and some functionality can be removed.

[0215] It should be noted that the above discussion has described a variety of different systems, components, logic, and interactions. It should be understood that any or all of such systems, components, logic, and interactions can be implemented by hardware items, such as processors, memories, or other processing components, including but not limited to artificial intelligence components (such as neural networks, some of which are described below) that perform the functionality associated with those systems, components, logic, or interactions. Furthermore, any or all of the systems, components, logic, and interactions can be implemented by software that is loaded into memory and subsequently executed by a processor or server or other computing component, as described below. Any or all of the systems, components, logic, and interactions can also be implemented by different combinations of hardware, software, firmware, and the like, some examples of which are described below. These are some examples of different structures that can be used to implement any or all of the systems, components, logic, and interactions described above. Other structures can also be used.

[0216] Figure 2 is a block diagram of an agricultural harvester 600, which can be similar to the agricultural harvester 100 shown in Figure 2 The agricultural harvester 600 communicates with elements in the remote server architecture 500. In some examples, the remote server architecture 500 can provide computing, software, data access, and storage services that do not require end users to be aware of the physical location or configuration of the system that delivers the services. In various examples, the remote server can deliver services over a wide area network, such as the Internet, using appropriate protocols. For example, the remote server can deliver an application over a wide area network, and the application can be accessed through a web browser or any other computing component. Figure 14 The software or components shown in

[0217] In Figure 2In the examples shown in FIG. 1, some items are similar to Figure 14 In the examples shown in FIG. 1, some items are similar to Figure 14 It is specifically shown that the prediction model generator 210 or the prediction map generator 212 or both can be located at a server location 502 remote from the agricultural harvester 600. Thus, in Figure 14 In the examples shown in FIG. 1, the agricultural harvester 600 accesses the system through the remote server location 502.

[0218] Figure 14 Another example of a remote server architecture is also depicted. Figure 2 It is shown that Figure 2 Some of the elements of FIG. 1 can be arranged at the remote server location 502, while other elements can be located elsewhere. As an example, the data store 202 can be placed at a location separate from the location 502 and accessed via a remote server at the location 502. Regardless of where these elements are located, these elements can be accessed by the agricultural harvester 600 directly through a network such as a wide area network or a local area network, these elements can be hosted by a service at a remote site, or these elements can be provided as a service or accessed by a connectivity service that resides at a remote location. Furthermore, data can be stored at any location and the stored data can be accessed or forwarded to an operator, user, or system by the operator, user, or system. For example, a physical carrier wave can be used instead of or in addition to an electromagnetic wave carrier. In some examples, another machine such as a fuel truck or other mobile machine or vehicle can have an automatic, semi-automatic, or manual information collection system in the event of poor or non-existent wireless telecommunication service coverage. The information collection system collects information from the combine harvester 600 using any type of temporary ad hoc wireless connection when the combine harvester 600 is in proximity to the machine containing the information collection system, such as a fuel truck, before the combine harvester 600 is refueled. The collected information can then be forwarded to another network when the machine containing the received information reaches a location where wireless telecommunication service coverage or other wireless coverage is available. For example, the fuel truck can enter an area with wireless communication coverage when the fuel truck travels to a location to refuel other machines or at a main fuel storage location. All of these architectures are contemplated herein. Furthermore, information can be stored on the agricultural harvester 600 until the agricultural harvester 600 enters an area with wireless communication coverage. The agricultural harvester 600 itself can transmit the information to another network.

[0219] It will also be noted that Figure 15The elements or portions thereof of the system can be provided on a variety of different devices. One or more of these devices can include an on-board computer, an electronic control unit, a display unit, a server, a desktop computer, a laptop computer, a tablet computer, or other mobile device such as a palmtop computer, a cellular telephone, a smart phone, a multimedia player, a personal digital assistant, and the like.

[0220] In some examples, the remote server architecture 500 can include network security measures. Without limitation, these measures include encryption of data on storage devices, encryption of data sent between network nodes, authentication of personnel or processes accessing data, and the use of ledgers to record metadata, data, data transfers, data access, and data transformations. In some examples, the ledgers can be distributed and immutable (e.g., implemented as blockchains).

[0221] Figures 16-17 is a simplified block diagram of one illustrative example of a handheld or mobile computing device that can be used as a handheld device 16 of a user or customer in which the present system (or a portion thereof) can be deployed. For example, the mobile device can be deployed in an operator's cab of an agricultural harvester 100 for use in generating, processing, or displaying the graphs discussed above. Figure 15 is an example of a handheld or mobile device.

[0222] Figure 2 An overall block diagram of components of a client device 16 is provided that can run some of the components shown in Figure 16 FIG. 1, interact with them, or both. In the device 16, a communication link 13 is provided that allows the handheld device to communicate with other computing devices, and in some examples a channel is provided for automatically receiving information (e.g., by scanning). Examples of the communication link 13 include allowing communication over one or more communication protocols, such as wireless services for providing cellular access to a network, and protocols for providing local wireless connections to a network.

[0223] In other examples, the applications can be received on a removable Secure Digital (SD) card that is connected to an interface 15. The interface 15 and the communication link 13 are in communication with a processor 17 (which can also implement the processor or server from other figures) along a bus 19 that is also connected to a memory 21 and input / output (I / O) components 23, as well as a clock 25 and a location system 27.

[0224] In one example, I / O components 23 are provided to facilitate input and output operations. Various examples of I / O components 23 in device 16 may include input components (such as buttons, touch sensors, optical sensors, microphones, touchscreens, proximity sensors, accelerometers, orientation sensors) and output components (such as display devices, speaker and / or printer ports). Other I / O components 23 may also be used.

[0225] Clock 25 schematically includes a real-time clock component that outputs the time and date. Schematically, it may also provide timing functions for processor 17.

[0226] Location system 27 schematically includes components that output the current geographic location of device 16. This may include, for example, a global positioning system (GPS) receiver, a LORAN system, a dead reckoning system, a cellular triangulation system, or other positioning systems. Location system 27 may also include, for example, mapping or navigation software that generates desired maps, navigation routes, and other geographic functions.

[0227] Memory 21 stores operating system 29, network settings 31, applications 33, application configuration settings 35, data storage device 37, communication driver 39, and communication configuration settings 41. Memory 21 may include all types of tangible, volatile, and non-volatile computer-readable storage devices. Memory 21 may also include computer storage media (described below). Memory 21 stores computer-readable instructions that, when executed by processor 17, cause the processor to perform computer-implemented steps or functions. Processor 17 may also be activated by other components to facilitate their functionality.

[0228] Figure 16 The illustration shows an example where device 16 is a tablet computer 600. Figure 17 In the diagram, computer 601 is shown as having a user interface display screen 602. Screen 602 can be a touchscreen or a pen-enabled interface that receives input from a pen or stylus. Tablet 600 can also utilize an on-screen virtual keyboard. Of course, computer 601 can also be attached to a keyboard or other user input device, for example, via a suitable attachment structure (such as a wireless link or USB port). Computer 601 can also schematically receive voice input.

[0229] Figure 16 Similar to Figure 18, except that the device is a smartphone 71. The smartphone 71 has a touch-sensitive display 73 that displays icons or widgets or other user input mechanisms 75. The mechanisms 75 can be used by a user to run applications, make calls, perform data transfer operations, etc. Generally speaking, the smartphone 71 builds on the mobile operating system and provides more advanced computing capability and connectivity than a feature phone.

[0230] Note that other forms of the device 16 are possible.

[0231] Figure 2 is one example of a computing environment in which elements of Figure 18 may be deployed. Referring to Figure 2 , an example system for implementing some embodiments includes a computing device in the form of a computer 810 programmed to operate as discussed above. The components of computer 810 can include, but are not limited to, a processing unit 820 (which can include a processor or server from the previous figures), a system memory 830, and a system bus 821 that couples various system components including the system memory to the processing unit 820. The system bus 821 can be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures. The memory and programs described with respect to Figure 18 may be deployed in corresponding portions of Figure 18 .

[0232] The computer 810 typically includes a variety of computer readable media. Computer readable media can be any available media that can be accessed by computer 810 and includes both volatile and nonvolatile media, removable and non-removable media. By way of example, and not limitation, computer readable media can comprise computer storage media and communication media. Computer storage media is different from, and does not include, a modulated data signal or carrier wave. It includes hardware storage media including both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by the computer 810. Communication media can embody computer readable instructions, data structures, program modules or other data in a modulated data signal or carrier wave. The term "modulated data signal" means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal.

[0233] System memory 830 includes computer storage media in the form of volatile and / or nonvolatile memory such as read only memory (ROM) 831 and random access memory (RAM) 832. A basic input / output system 833 (BIOS), containing the basic routines that help to transfer information between elements within computer 810, such as during start-up, is typically stored in ROM 831. RAM 832 typically contains data and / or program modules that are immediately accessible to and / or presently being operated on by processing unit 820. By way of example, and not limitation, Figure 8 illustrates operating system 834, application programs 835, other program modules 836, and program data 837. Figure 18

[0234] Computer 810 can also include other removable / non-removable volatile / nonvolatile computer storage media. By way of example only, Figure 18 Hard disk drive 841 reads from or writes to non-removable, nonvolatile magnetic media (e.g., a not shown), optical disk drive 855 reads from or writes to a removable, nonvolatile optical disk 856 such as a DVD-ROM or other optical media. Other removable / non-removable, volatile / nonvolatile computer storage media that can be used with computer 810, such as solid state memory, flash memory, magnetic tape, or optical storage are exemplified by way of example, and not limitation. It is to be appreciated that the hard disk drive 841, and the optical disk drive 855 are typically

[0235] Alternatively, or in addition, the functionality described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Application-specific Integrated Circuits (ASICs), Application-specific Standard Products (e.g., ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0236] The drives and their associated computer storage media discussed above and illustrated in Figure 18 Figure 8, provide storage of computer readable instructions, data structures, program modules and other data for the computer 810. In Figure 8, for example, hard disk drive 841 is illustrated as storing operating system 844, application programs 845, other program modules 846, and program data 847. Figure 18 ​In this diagram, hard disk drive 841 is shown storing operating system 844, application programs 845, other program modules 846, and program data 847. Note that these components may be the same as or different from operating system 834, application programs 835, other program modules 836, and program data 837.

[0237] Users can input commands and information into computer 810 using input devices such as keyboard 862, microphone 863, and pointing devices 861 (such as mouse, trackball, or touchpad). Other input devices (not shown) may include joysticks, game controllers, satellite dishes, scanners, etc. These and other input devices are typically connected to processing unit 820 via user input interface 860 coupled to the system bus, but may also be connected via other interfaces and bus structures. Visual display 891 or other types of display devices are also connected to system bus 821 via an interface such as video interface 890. In addition to the monitor, the computer may also include other peripheral output devices, such as speakers 897 and printers 896, which can be connected via peripheral output interface 895.

[0238] Computer 810 operates in a networked environment using logical connections (such as Controller Area Network (CAN), Local Area Network (LAN), or Wide Area Network (WAN)) to one or more remote computers (such as remote computer 880).

[0239] When used in a LAN networking environment, computer 810 connects to LAN 871 via a network interface or adapter 870. When used in a WAN networking environment, computer 810 typically includes a modem 872 or other devices for establishing communication over a WAN 873 (such as the Internet). In a networking environment, program modules can be stored in remote memory storage devices. For example, ​ This demonstrates that remote application 885 can reside on remote computer 880.

[0240] It should also be noted that the different examples described in this article can be combined in different ways. That is, parts of one or more examples can be combined with parts of one or more other examples. All of this is considered in this article.

[0241] Example 1 is an agricultural operating machine, comprising:

[0242] A communication system that receives an information map including values ​​of weed characteristics corresponding to different geographical locations in a field;

[0243] A geolocation sensor that detects the geographical location of agricultural machinery;

[0244] a field sensor that detects a value of an agricultural property corresponding to a geographic location;

[0245] a predicted map generator that generates a functional predicted agricultural map of the field that maps predicted control values to different geographic locations in the field based on the value of the weed property in the information map and based on the value of the agricultural property;

[0246] a controllable subsystem; and

[0247] a control system that generates control signals to control the controllable subsystem based on the geographic location of the agricultural work machine and based on the control values in the functional predicted agricultural map.

[0248] Example 2 is the agricultural work machine of any or all preceding examples, wherein the predicted map generator comprises:

[0249] a predicted biomass map generator that generates a functional predicted biomass map that maps predicted biomass of the material to different geographic locations in the field.

[0250] Example 3 is the agricultural work machine of any or all preceding examples, wherein the control system comprises:

[0251] a feed rate controller that generates a feed rate control signal based on the detected geographic location and the functional predicted biomass map and controls the controllable subsystem based on the feed rate control signal to control a feed rate of the material through the agricultural work machine.

[0252] Example 4 is the agricultural work machine of any or all preceding examples, wherein the predicted map generator comprises:

[0253] a predicted machine speed map generator that generates a functional predicted machine speed map that maps predicted machine speed values to different geographic locations in the field.

[0254] Example 5 is the agricultural work machine of any or all preceding examples, wherein the control system comprises:

[0255] a setting controller that generates a speed control signal based on the detected geographic location and the functional predicted machine speed map and controls the controllable subsystem based on the speed control signal to control a speed of the agricultural work machine.

[0256] Example 6 is the agricultural work machine of any or all preceding examples, wherein the predicted map generator comprises:

[0257] a predicted operator command map generator that generates a functional predicted operator command map that maps predicted operator commands to different geographic locations in the field.

[0258] Example 7 is the agricultural work machine of any or all preceding examples, wherein the control system comprises:

[0259] a settings controller that generates an operator command control signal indicative of an operator command based on the detected geographic location and the functional predicted operator command map, and controls the controllable subsystem to perform the operator command based on the operator command control signal.

[0260] Example 8 is the agricultural work machine of any or all preceding examples, and further comprising:

[0261] a predictive model generator that generates a predictive agricultural model that models a relationship between a weed characteristic and an agricultural characteristic based on values of the weed characteristic in the prior information map at the geographic location and values of the agricultural characteristic sensed by the field sensor at the geographic location, wherein the predictive map generator generates the functional predicted agricultural map based on the values of the weed characteristic in the prior information map and based on the predictive agricultural model.

[0262] Example 9 is the agricultural work machine of any or all preceding examples, wherein the control system further comprises:

[0263] an operator interface controller that generates a user interface graphical representation of the functional predicted agricultural map that includes a field portion having a current location indicator indicative of a geographic location of the agricultural work machine on the field portion and a weed characteristic symbol indicative of a value at one or more geographic locations on the field portion.

[0264] Example 10 is the agricultural work machine of any or all preceding examples, wherein the operator interface controller generates the user interface graphical representation to include an interactive display portion that displays a sensed characteristic display indicative of the sensed agricultural characteristic, an interactive threshold display portion indicative of an action threshold, and an interactive action indicator indicative of a control action to be taken when the sensed agricultural characteristic satisfies the action threshold, the control system generating a control signal to control the controllable subsystem based on the control action.

[0265] Example 11 is a computer-implemented method of controlling an agricultural work machine, comprising

[0266] obtaining a prior information map comprising values of a weed characteristic corresponding to different geographic locations in a field;

[0267] detecting a geographic location of the agricultural work machine;

[0268] detecting, with an on-site sensor, a value of an agricultural characteristic corresponding to a geographic location;

[0269] generating, based on the value of the weed characteristic in the prior information map and based on the value of the agricultural characteristic, a functional predictive agricultural map that maps predictive control values to different geographic locations in the field; and

[0270] controlling the controllable subsystem based on the geographic location of the agricultural work machine and based on the control values in the functional predictive agricultural map.

[0271] Example 12 is the computer-implemented method of any or all preceding examples, wherein generating the functional predictive map comprises:

[0272] generating a functional predictive biomass map that maps predicted biomass of the material to different geographic locations in the field.

[0273] Example 13 is the computer-implemented method of any or all preceding examples, wherein controlling the controllable subsystem comprises:

[0274] generating a feed rate control signal based on the detected geographic location and the functional predictive biomass map; and

[0275] controlling the controllable subsystem based on the feed rate control signal to control a feed rate of the material by the agricultural work machine.

[0276] Example 14 is the computer-implemented method of any or all preceding examples, wherein generating the functional predictive map comprises:

[0277] generating a functional predictive machine speed map that maps predicted machine speed values to different geographic locations in the field.

[0278] Example 15 is the computer-implemented method of any or all preceding examples, wherein controlling the controllable subsystem comprises:

[0279] generating a speed control signal based on the detected geographic location and the functional predictive machine speed map; and

[0280] controlling the controllable subsystem based on the speed control signal to control a speed of the agricultural work machine.

[0281] Example 16 is the computer-implemented method of any or all preceding examples, wherein generating the functional predictive map comprises:

[0282] generating a functional predictive operator command map that maps predicted operator commands to different geographic locations in the field.

[0283] Example 17 is the computer-implemented method of any or all preceding examples, wherein controlling the controllable subsystem comprises:

[0284] generate an operator command control signal indicative of the operator command based on the detected geographic location and the function predicted operator command map; and

[0285] control the controllable subsystem to perform the operator command based on the operator command control signal.

[0286] Example 18 is a computer-implemented method of any or all of the preceding examples, and further comprising:

[0287] generate a predictive agricultural model modeling a relationship between the weed characteristic and the agricultural characteristic based on the value of the weed characteristic in the prior information map at the geographic location and the value of the agricultural characteristic sensed by the field sensor at the geographic location, wherein generating the function predicted agricultural map comprises generating the function predicted agricultural map based on the value of the weed characteristic in the prior information map and based on the predictive agricultural model.

[0288] Example 19 is an agricultural work machine comprising:

[0289] a communication system that receives a prior information map comprising values of a weed characteristic corresponding to different geographic locations in a field;

[0290] a geographic location sensor that detects a geographic location of the agricultural work machine;

[0291] a field sensor that detects a value of an agricultural characteristic corresponding to the geographic location;

[0292] a predictive model generator that generates a predictive agricultural model modeling a relationship between the weed characteristic and the agricultural characteristic based on the value of the weed characteristic in the prior information map at the geographic location and the value of the agricultural characteristic sensed by the field sensor at the geographic location;

[0293] a predictive map generator that generates a function predicted agricultural map of the field mapping predicted control values to different geographic locations in the field based on the value of the weed characteristic in the prior information map and based on the predictive agricultural model;

[0294] a controllable subsystem; and

[0295] a control system that generates a control signal to control the controllable subsystem based on the geographic location of the agricultural work machine and based on the control values in the function predicted agricultural map.

[0296] Example 20 is the agricultural work machine of any or all of the preceding examples, wherein the control system comprises at least one of:

[0297] a feed rate controller that generates a feed rate control signal based on the detected geographic location and the function-predicted agricultural map and controls the controllable subsystem based on the feed rate control signal to control a feed rate of material through the agricultural work machine;

[0298] a speed controller that generates a speed control signal based on the detected geographic location and the function-predicted agricultural map and controls the controllable subsystem based on the speed control signal to control a speed of the agricultural work machine; and

[0299] an operator command controller that generates an operator command control signal indicative of an operator command based on the detected geographic location and the function-predicted agricultural map and controls the controllable subsystem based on the operator command control signal to execute the operator command.

[0300] Another example 1 is an agricultural work machine (100) comprising:

[0301] a communication system (206) that receives an a priori information map (258) comprising values of a first agricultural property corresponding to different geographic locations in a field;

[0302] a geographic location sensor (204) that detects a geographic location of the agricultural work machine (100);

[0303] a field sensor (208) that detects a value of a second agricultural property corresponding to the geographic location;

[0304] a predictive model generator (210) that generates a predictive agricultural model that models a relationship between the first agricultural property and the second agricultural property based on the value of the first agricultural property at the geographic location in the a priori information map (258) and the value of the second agricultural property sensed by the field sensor (208) at the geographic location; and

[0305] a predictive map generator (212) that generates a function-predicted agricultural map of the field that maps predicted values of the second agricultural property to the different geographic locations in the field based on the values of the first agricultural property in the a priori information map (258) and based on the predictive agricultural model.

[0306] Another example 2 is the agricultural work machine of any or all preceding examples, wherein the prediction map generator configures the functional predicted agricultural map for use by a control system that generates control signals for controlling controllable subsystems on the agricultural work machine based on the functional predicted agricultural map.

[0307] Another example 3 is the agricultural work machine of any or all preceding examples, wherein the field sensor on the agricultural work machine is configured to detect a weed characteristic corresponding to the geographic location as the value of the second agricultural characteristic.

[0308] Another example 4 is the agricultural work machine of any or all preceding examples, the field sensor comprising:

[0309] an image detector configured to detect an image indicative of the weed characteristic.

[0310] Another example 5 is the agricultural work machine of any or all preceding examples, wherein the image detector is oriented to detect images in a vicinity of the agricultural work machine, and

[0311] the image detector further comprises:

[0312] an image processing system configured to process the image to identify the weed characteristic in the image.

[0313] Another example 6 is the agricultural work machine of any or all preceding examples, wherein:

[0314] the field sensor generates a sensor signal indicative of the weed characteristic, and

[0315] the field sensor further comprises:

[0316] a processing system that receives the sensor signal and is configured to identify a weed intensity value indicative of a weed intensity at the geographic location as the weed characteristic.

[0317] Another example 7 is the agricultural work machine of any or all preceding examples, wherein:

[0318] the field sensor generates a sensor signal indicative of the weed characteristic, and

[0319] the field sensor further comprises:

[0320] a processing system that receives the sensor signal and is configured to identify a weed type characteristic value indicative of a weed type at the geographic location as the weed characteristic.

[0321] Another example 8 is the agricultural work machine of any or all preceding examples, wherein,

[0322] the image detector is configured to capture images including at least one weed seed processed by the agricultural work machine, and

[0323] wherein the image processing system is configured to identify, based on the images of the at least one weed seed, a weed intensity value indicative of a weed intensity at the geographic location, or to identify a weed type value indicative of a weed type at the geographic location.

[0324] Another example 9 is the agricultural work machine of any or all preceding examples, wherein,

[0325] the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the different geographic locations in the field, and

[0326] wherein the predictive model generator is configured to identify, based on the weed property values detected at the geographic locations and the vegetation index values at the geographic locations in the vegetation index map, a relationship between the vegetation index values and the weed property, the predictive agricultural model being configured to receive vegetation index values as model input and generate weed property values as model output based on the identified relationship.

[0327] Another example 10 is the agricultural work machine of any or all preceding examples, wherein,

[0328] the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the different geographic locations in the field, and

[0329] wherein the predictive model generator is configured to identify, based on the weed intensity values detected at the geographic locations and the vegetation index values at the geographic locations in the vegetation index map, a relationship between the vegetation index values and the weed intensity property, the predictive agricultural model being configured to receive vegetation index values as model input and generate weed intensity values as model output based on the identified relationship.

[0330] Another example 11 is the agricultural work machine of any or all preceding examples, wherein,

[0331] the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the field, and

[0332] wherein the prediction model generator is configured to identify a relationship between the vegetation index values and the weed type characteristics based on the weed type values detected at the geographic locations and the vegetation index values at the geographic locations in the vegetation index map, the predictive agriculture model is configured to receive a vegetation index value as a model input and generate a weed type value as a model output based on the identified relationship.

[0333] Another example 12 is a computer-implemented method of generating a functional predictive agriculture map, comprising:

[0334] receiving, at an agricultural work machine (100), a prior information map (258) indicating values of a first agricultural characteristic corresponding to different geographic locations in a field;

[0335] detecting a geographic location of the agricultural work machine (100);

[0336] detecting, with an in-field sensor (208), a value of a second agricultural characteristic corresponding to the geographic location;

[0337] generating a predictive agriculture model modeling a relationship between the first agricultural characteristic and the second agricultural characteristic; and

[0338] controlling a predictive map generator (212) to generate the functional predictive agriculture map for the field mapping predicted values of the second agricultural characteristic to the different locations in the field based on the values of the first agricultural characteristic in the prior information map (258) and the predictive agriculture model.

[0339] Another example 13 is the computer-implemented method of any or all preceding examples, further comprising:

[0340] configuring the functional predictive agriculture map for a control system that generates control signals for controlling controllable subsystems on the agricultural work machine based on the functional predictive agriculture map.

[0341] Another example 14 is the computer-implemented method of any or all preceding examples, wherein detecting the value of the second agricultural characteristic with an in-field sensor comprises detecting a weed characteristic corresponding to the geographic location, and

[0342] wherein receiving the prior information map comprises receiving a prior vegetation index map mapping vegetation index values as the first agricultural characteristic to the different geographic locations in the field.

[0343] Another example 15 is an agricultural work machine (100), comprising:

[0344] a communication system (206) that receives an a priori vegetation index map (332) indicative of vegetation index values corresponding to different geographic locations in a field;

[0345] a geographic location sensor (204) that detects a geographic location of the agricultural work machine (100);

[0346] a field sensor (208) that detects a weed characteristic value of a weed characteristic corresponding to the geographic location;

[0347] a predictive model generator (210) that generates a predictive weed model modeling a relationship between the vegetation index values and the weed characteristic based on a vegetation index value at the geographic location in the a priori vegetation index map (332) and the weed characteristic value of the weed characteristic sensed by the field sensor (208) at the geographic location; and

[0348] a predictive map generator (212) that generates a functional predictive weed map of the field mapping predictive weed characteristic values to the different locations in the field based on the vegetation index values in the a priori vegetation index map (332) and based on the predictive weed model.

[0349] Although the subject matter has been described in language specific to structural features or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. An agricultural work machine (100) comprising: a communication system (206) that receives a prior information map (258) comprising values of a first agricultural property corresponding to different geographic locations in a field; a geographic location sensor (204) that detects a geographic location of the agricultural work machine (100); a field sensor (208) that detects a value of a second agricultural property corresponding to the geographic location; a predictive model generator (210) that generates a predictive agricultural model that models a relationship between the first agricultural property and the second agricultural property based on the value of the first agricultural property at the geographic location in the prior information map (258) and the value of the second agricultural property sensed by the field sensor (208) at the geographic location; and a predictive map generator (212) that generates a functional predictive agricultural map of the field that maps predicted values of the second agricultural property to the different geographic locations in the field based on the values of the first agricultural property in the prior information map (258) and based on the predictive agricultural model.

2. The agricultural work machine of claim 1, wherein, The predictive map generator configures the functional predictive agricultural map for use by a control system that generates control signals for controlling controllable subsystems on the agricultural work machine based on the functional predictive agricultural map.

3. The agricultural work machine of claim 1, wherein, The field sensor on the agricultural work machine is configured to detect a value of a weed property corresponding to the geographic location as the value of the second agricultural property.

4. The agricultural work machine of claim 3, wherein, The field sensor comprises: an image detector configured to detect an image indicative of the weed property.

5. The agricultural work machine of claim 4, wherein, The image detector is oriented to detect an image of a vicinity of the agricultural work machine, and The image detector further comprises: an image processing system configured to process the image to identify the value of the weed property in the image.

6. The agricultural work machine of claim 3, wherein, the field sensor generates a sensor signal indicative of the value of the weed property, and the field sensor further comprises: a processing system that receives the sensor signal and is configured to identify a value of a weed intensity indicative of a weed intensity at the geographic location as the value of the weed property.

7. The agricultural work machine of claim 3, wherein, the field sensor generates a sensor signal indicative of the value of the weed property, and the field sensor further comprises: a processing system that receives the sensor signal and is configured to identify a value of a weed type indicative of a weed type at the geographic location as the value of the weed property.

8. The agricultural work machine of claim 5, wherein, the image detector is configured to capture images, the images including at least one weed seed processed by the agricultural work machine, and wherein the image processing system is configured to identify, based on the images of the at least one weed seed, a value of weed intensity indicative of a weed intensity at the geographic location, or to identify a value of weed type indicative of a weed type at the geographic location.

9. The agricultural work machine of claim 3, wherein the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the different geographic locations in the field, and wherein the predictive model generator is configured to identify, based on the value of the weed property detected at the geographic location and the vegetation index value at the geographic location in the vegetation index map, a relationship between the vegetation index value and the weed property, the predictive agricultural model being configured to receive a vegetation index value as a model input and generate a value of the weed property as a model output based on the identified relationship.

10. The agricultural work machine of claim 6, wherein the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the different geographic locations in the field, and wherein the predictive model generator is configured to identify, based on the value of the weed intensity detected at the geographic location and the vegetation index value at the geographic location in the vegetation index map, a relationship between the vegetation index value and weed intensity, the predictive agricultural model being configured to receive a vegetation index value as a model input and generate a value of weed intensity as a model output based on the identified relationship.

11. The agricultural work machine of claim 7, wherein the prior information map comprises a prior vegetation index map that maps vegetation index values as the first agricultural property to the field, and wherein the predictive model generator is configured to identify, based on the value of the weed type detected at the geographic location and the vegetation index value at the geographic location in the vegetation index map, a relationship between the vegetation index value and the weed type, the predictive agricultural model being configured to receive a vegetation index value as a model input and generate a value of weed type as a model output based on the identified relationship.

12. A computer-implemented method of generating a functional predictive agriculture map, comprising: receiving, at an agricultural work machine (100), a prior information map (258) indicative of values of a first agricultural property corresponding to different geographic locations in a field; detecting a geographic location of the agricultural work machine (100); detecting, with an in-field sensor (208), a value of a second agricultural property corresponding to the geographic location; generating a predictive agricultural model that models a relationship between the first agricultural property and the second agricultural property; and a control prediction map generator (212) is controlled to generate, based on the values of the first agricultural property in the prior information map (258) and the predictive agricultural model, a functional predictive agricultural map of the field that maps predicted values of the second agricultural property to the different geographical locations in the field.

13. The computer-implemented method of claim 12, further comprising: configuring the functional predictive agricultural map for a control system that generates control signals for controlling controllable subsystems on the agricultural work machine based on the functional predictive agricultural map.

14. The computer-implemented method of claim 12, wherein, detecting, with the field sensor, values of the second agricultural property includes detecting a weed property corresponding to the geographical location, and wherein receiving the prior information map includes receiving a prior vegetation index map that maps vegetation index values as the first agricultural property to the different geographical locations in the field.

15. An agricultural work machine (100), comprising: a communication system (206) that receives a prior vegetation index map (332) that indicates vegetation index values corresponding to different geographical locations in a field; a geographical location sensor (204) that detects a geographical location of the agricultural work machine (100); a field sensor (208) that detects a weed property value of a weed property corresponding to the geographical location; a predictive model generator (210) that generates, based on a vegetation index value at the geographical location in the prior vegetation index map (332) and the weed property value of the weed property sensed by the field sensor (208) at the geographical location, a predictive weed model that models a relationship between the vegetation index value and the weed property; and a prediction map generator (212) that generates, based on the vegetation index value in the prior vegetation index map (332) and based on the predictive weed model, a functional predictive weed map of the field that maps predicted weed property values to the different geographical locations in the field.

Images