Agricultural baler, as well as method and device for determining the bale weight of a harvested crop bale formed therefrom.

Abstract

The present invention relates to determining the bale weight of a crop bale formed by an agricultural baler. The invention relates, on the one hand, to a method and a device for determining the bale weight using various sensor signals from sensors located on the baler, and on the other hand, to the baler itself. According to the invention, the baler has sensors for detecting the moisture content of the crop flow taken in by the baler, for detecting the bale density of the crop bale in the baling chamber, and for detecting the bale diameter.

Description

[0001] The present invention relates to determining the bale weight of a harvested crop bale formed by an agricultural baler. The invention relates, on the one hand, to a method and a device for determining the bale weight using various sensor signals from sensors located on the baler, and on the other hand, to the baler itself.

[0002] Determining the weight of harvested crop bales can be useful for various reasons, for example for recording the yield per unit area or for controlling settings on the wrapping device or the baler.

[0003] Up to now, the weight measurement of harvested crop bales has mostly been carried out on the wrapping table, which is provided behind the bale forming chamber and usually has a weight sensor such as a force measuring bolt, cf. for example NL 20 10 169 C, EP 20 00 785 B1, EP 10 34 695 B1, EP 27 45 675 B1 or EP 40 56 024 A1.

[0004] Weighing systems can generally operate with various types of sensors and may include, for example, pressure sensors or load cells that measure the pressure exerted by the bale's weight on the baler's or wrapping table's support arms. Alternatively, accelerometers are used to measure the change in acceleration caused by picking up the bale. Strain gauges are also employed, detecting the deformation of support arms or other chassis elements to determine the bale's weight.

[0005] However, such weighing systems are often sensitive and always expensive. At the same time, measurement errors can occur, for example, if the baler or the attached wrapping table is tilted or excessively inclined when driving on slopes, at least if no expensive compensation measures are taken to ensure accurate measurements even on an inclined surface.

[0006] Based on this, the present invention aims to provide an improved baler of the aforementioned type, as well as improved methods and devices for determining the bale weight of the crop bales formed by the baler, avoiding the disadvantages of the prior art and advantageously developing the latter further. In particular, it should enable an accurate determination of the bale weight using the simple sensors normally present on a baler, without requiring expensive weighing systems or costly compensation measures for such weighing.

[0007] According to the invention, the aforementioned problem is solved by a method according to claim 1, a device according to claim 12, and an agricultural baler according to claim 14. Preferred embodiments of the invention are the subject of the dependent claims.

[0008] It is therefore proposed that, during regular operation of the baler, the bale weight of the harvested crop bales be determined without actual weighing and thus without the aid of a weighing system. Instead, the bale weight is estimated based on meaningful parameters provided by the sensors already present on the baler for other purposes. To ensure sufficient accuracy in this estimation, the aforementioned parameters, in the form of sensor signals, are recorded and analyzed during a training phase, along with the bale weights actually measured during this phase. This analysis aims to determine the functional influence of the sensor-acquired parameters on the actual bale weight.Based on the relationship between the sensor-detected parameters and the actual bale weights known from the training phase, the bale weight can then be determined in the operating phase without a signal from a weighing system, solely based on the sensor-detected parameters and the known relationship.

[0009] According to the invention, the baler has sensors for detecting the moisture content of the crop flow taken in by the baler, for detecting the bale density of the crop bale in the baling chamber, and for detecting the bale diameter. Such sensors are already present on many balers to perform other control tasks, for example, to adjust the tension of the variable belts. During a training phase, the moisture content of the crop flow, the bale density of the crop bale formed in the baling chamber of the baler, and the bale diameter are detected for several baling operations using the aforementioned sensors. Additionally, the weight of each formed crop bale is also measured during the training phase using a weighing system.In a training memory, the data sets from the training phase, which include the recorded moisture content, bale density, and bale diameter along with the corresponding measured bale weights, are stored. The stored data sets are then analyzed by an evaluation unit to determine correlations between the measured bale weights and the recorded moisture content, bale density, and bale diameter.In an operational phase following the training phase, the aforementioned sensors are used to record the current moisture content of the incoming crop flow, the current bale density of the crop bale formed in the baling chamber of the baler, and the current bale diameter for each bale formation. The evaluation unit that analyzed the training data, or another evaluation unit, then uses the relationships previously determined in the training phase to calculate the bale weight of the currently formed bale from the currently recorded moisture content, the currently recorded bale density, and the currently recorded bale diameter, without a measurement signal from a weighing system.

[0010] The proposed solution eliminates the need for an expensive weighing system during regular operation. In particular, no special compensation measures are required to obtain precise bale weights, even when driving on slopes or when the baler is generally tilted. Furthermore, it is advantageous that only the existing sensors on the baler are used, eliminating the need for additional sensors. This allows for a more cost-effective and robust baler design, as well as the elimination of weighing systems with integrated measuring equipment. Simultaneously, by estimating the bale weight based on the measured moisture content of the incoming crop, the current bale density, and the current diameter, a future-oriented prediction of the final bale weight can be made even during the baling process. This enables rapid decision-making, such as...Predicting bale size at a certain distance or at the end of the tramline, or checking bale weights during operation.

[0011] Surprisingly, a very high accuracy can be achieved in determining the bale weight, with a deviation from the actual bale weight of less than 10% or even less than 5%.

[0012] In a further development of the invention, the moisture content of the harvested crop, the bale density, and the bale diameter are determined by the sensors in real time. The real-time sensors can provide these parameters—moisture content, bale density, and bale diameter—in real time both during the training phase and in the subsequent, regular operating phase. This makes it possible to store not only individual values ​​of these parameters in the training memory, but also their progression over each bale-forming process. This allows the evaluation unit to assess the correlation between the progression of moisture content, bale density, and bale diameter over a bale-forming process and the resulting bale weight that was actually measured.This is particularly helpful because the moisture content of the incoming crop stream can vary during the baling process. For example, it might be higher in a tramline at the edge of a forest and lower in a tramline in a sunny section of a field. Similarly, the bale density of the bale being formed in the chamber can also vary throughout the process. For instance, it might be higher when a more voluminous layer of crop is picked up by the baler's intake device and fed into the baling pre-chamber, and lower when the crop thins out. Likewise, the bale diameter can increase not only linearly but also at different rates during the baling process.

[0013] Preferably, the sensor signal profiles indicating moisture content, bale density, and bale diameter are linked to the measured, actual bale weight of each bale forming process and stored in the training memory. This training memory contains various data sets, each comprising the aforementioned signal profiles and the corresponding measured bale weight. A data set can therefore, in particular, include the moisture content, bale density, and developing bale diameter profiles over a bale forming process, along with the bale weight measured for that process. The evaluation unit can then analyze these data sets to determine the correlation between the actual bale weight and the aforementioned profiles.

[0014] If the aforementioned parameters are recorded in real time, the evaluation unit can, for example, consider the slope and / or curvature of the signal waveforms and determine their effects on the relationship to be determined with the measured bale weight. For instance, after a relatively high moisture content at the beginning of the baling process, a steeper drop in moisture content may occur. From this, the evaluation unit can conclude that the increased moisture content of the crop flow was caused by a specific feature, such as a tramline section in a moisture-deprived area, and will therefore only occur partially. This will result in a significantly lower overall moisture content, which in turn will lead to only a smaller increase in bale weight due to a smaller proportion of crop with higher moisture content.

[0015] Alternatively or additionally to evaluating the slope and curvature of the signal waveforms, the evaluation unit can also integrate the signal waveforms to determine, correct, or estimate the total moisture content, total bale density, and total bale diameter. In particular, the evaluation unit can also be configured to correlate signal waveforms to determine which moisture levels and / or densities occur in which partial areas of the crop bale being formed. If the evaluation unit knows the current diameter of the bale forming in the chamber, it can estimate in which section a currently entering portion of the crop layer and its moisture content is entering or being formed into the bale.For example, when bale diameters are still very small at the beginning of the forming process, the harvested crop with higher moisture content is incorporated into the core of the harvested crop bale, so to speak, and thus only into a smaller part, so that the evaluation device can deduce what proportion of the harvested crop with higher moisture content will be in the finished bale.

[0016] In particular, the aforementioned evaluation unit can include an AI module for determining the relationship between the bale weights measured during the training phase and the moisture content, bale density, and bale diameter recorded during the same phase. Such an AI module can, for example, include a regression analysis module to estimate the relationship between the parameters recorded by the sensors and the bale weights measured during the training phase, and potentially also between the dependencies of the sensor-recorded parameters themselves. For instance, such a regression analysis module can determine a functional relationship between the recorded parameters moisture content, bale density, and bale diameter on the one hand, and bale weight on the other, or generate a curve via regression that characterizes the dependence of bale weight on the aforementioned sensor-recorded parameters.

[0017] In particular, the aforementioned AI component can also include or have access to a neural network, so that parameters recorded during the training phase, especially the recorded moisture levels, bale densities, and bale diameters, as well as the bale weights measured by the weighing system, are fed into the neural network in order to determine the aforementioned relationships between the sensor-recorded parameters and the measured bale weight, and to derive a rule for estimating the bale weight from the aforementioned parameters alone, i.e., without the aid of a weighing system, for the later operational phase.

[0018] Such a neural network, or more generally the aforementioned evaluation device, can, for example, use a weighting module to assign a greater or lesser weight to one or more of the sensor-detected parameters or an anomaly—for example, a deviation of a sensor-detected parameter from an average value from other training sessions—than to another sensor-detected parameter or its deviation from an average value. For instance, the weighting can be adjusted so that, for the three sensor-detected parameters mentioned—moisture, bale density, and bale diameter—a greater weight is assigned to a larger deviation of bale density from the average than to a larger deviation of bale diameter from the average, and the bale weight estimate is corrected less or more strongly, according to the weighting.

[0019] To further increase the accuracy of bale weight estimation, additional parameters can be considered and, if necessary, measured by sensors. Advantageously, at least the number of cutting knives of a cutting unit that cuts the crop flow on its way to the bale forming chamber is taken into account. For this purpose, several bale forming cycles are advantageously performed during the training phase, using different numbers of knives. The respective number of knives is stored with the actual bale weight in the corresponding data sets, so that each data set includes the aforementioned parameters moisture content, bale density, and bale diameter, as well as the number of cutting knives used in the bale forming process, and also the actual bale weight measured by the weighing device.The evaluation unit, in particular the aforementioned AI component, evaluates these data sets from the training phase in order to derive a correlation between the number of cutting knives and the resulting bale weight, and then to take the number of cutting knives into account when estimating the bale weight in the operating phase.

[0020] In particular, a higher number of cutting knives, which regularly leads to a shorter cutting length of the cut crop stalks or straws, can lead to a higher bale density and thus also to a higher final weight of the crop bale, while conversely, a cutting unit that works with fewer cutting knives allows longer crop material to enter the bale forming chamber, so a lower bale weight can be expected.

[0021] The number of cutting blades can be detected by sensors, for example, by monitoring the occupancy of the available mounting positions for cutting blades. Alternatively or additionally, the number of cutting blades can also be entered manually via an input device or read from a setup status stack.

[0022] To achieve greater accuracy in estimating bale weight, the crop type can also be considered as an additional parameter. For example, a distinction can be made between at least two or even more crop types such as alfalfa, green rye, ryegrass, oat straw, barley straw, wheat straw, or mixed crops. The crop type can be entered manually, read from or imported from a field map, or automatically determined by sensors, such as a camera system with image processing. The camera system can optically capture the crop during mowing or directly while collecting the crop stream, allowing the image processing unit to determine the crop type by comparison with reference images.Alternatively or additionally, a crop type determination device can also be designed in other ways, for example by having access to a field record or accessing records from other field processing machines such as a combine harvester to obtain information about the crop type.

[0023] The crop type is advantageously recorded or entered during the training phase for the respective bale forming processes, or determined in the manner mentioned above, and stored in the training memory with the data records, so that the respective data records also include the crop type that characterizes the respective crop bale, in addition to the parameters mentioned above.

[0024] The aforementioned evaluation device, in particular the aforementioned AI component, also takes into account the additional information on the crop type in order to determine a relationship between crop type and bale weight from the training data sets, in conjunction with the aforementioned additional parameters bale diameter, bale density and moisture, and the bale weight, so that when estimating the bale weight in the operational phase, the respective crop processed or constituting the bale can also be used for estimating the bale weight.

[0025] To determine the moisture content of the harvested crop as precisely as possible, which is crucial for estimating bale weight, especially online and continuously with fast and / or varying crop flows, a moisture sensor can be provided on the baler, particularly in the area of ​​its intake device, and / or optionally, a moisture sensor can also be located in the bale forming chamber. Regardless of its positioning, the moisture sensor can advantageously incorporate a conductivity and / or resistance measuring module to determine the moisture content by measuring conductivity or resistance. Alternatively or additionally, the moisture sensor can also be designed to measure capacitively, in the near-infrared range, and / or using microwaves.

[0026] Advantageously, the humidity sensor can also include several measuring modules that work with different measuring principles, for example a conductance and / or resistance measuring module as well as a capacitive measuring module, in order to be able to determine the humidity precisely.

[0027] To precisely measure or record bale density, a parameter relevant to bale weight, a density sensor can be used to indirectly measure or determine the bale density. Specifically, bale density can be derived from the compression pressure in the bale forming chamber or calculated from the compression pressure, in which case the bale density sensor is designed as a compression pressure sensor.

[0028] Such a pressure sensor can be configured in particular to detect the tension force and / or tensile force of one or more belts encircling the bale in the bale forming chamber, since the belt tension force is a reliable measure of bale density.

[0029] For example, in a pressure actuator, particularly in a hydraulic cylinder or pressure supply system, the pressure of the pressure medium can be measured by a pressure sensor, which is then used as a measure of bale density. This pressure medium actuator can be a tensioning actuator that tensions one or more belts in the bale forming chamber.

[0030] Alternatively or in addition to a belt tension sensor, the bale density sensor can also have a bearing reaction force sensor, which measures the bearing reaction force on a deflection roller acted upon by the bale in the bale forming chamber.

[0031] Advantageously, the bale density sensor system mentioned above can include several sensors of the type mentioned above in order to obtain signals from different sources that characterize the bale density, so that a bale density signal value can be derived from the various individual signals.

[0032] The bale diameter can also be determined by the sensor in various ways, with a bale diameter sensor advantageously being designed to operate without contact. In particular, the bale diameter sensor can be designed as a distance sensor that determines or detects the distance of the bale surface from a fixed or known mounting point. For example, an ultrasonic sensor can be provided to determine the distance of the bale's circumferential surface in the bale forming chamber from a fixed mounting point.

[0033] Alternatively or additionally, the bale diameter sensor can also include an angle sensor to determine the angle of a bale surface to a predetermined reference surface or orientation. The bale diameter can then be calculated from this determined angle.

[0034] The invention is explained in more detail below with reference to a preferred embodiment and the accompanying drawings. The drawings show: Fig. 1: A side view of a baling press according to an advantageous embodiment of the invention, and Fig. 2: A side view of the interior of the baler made of Fig. 1, which shows the bale forming chamber as well as the conveying rotor and the crop flow conveyed into the bale forming chamber along with the sensors for recording operating and crop parameters.

[0035] How Fig. Figure 1 shows that the baler 1 can be designed as an attachment that can be mounted on a tractor not shown, for example by means of a drawbar 18 which is connected to a machine frame 19 which is supported on the ground by a chassis 20.

[0036] A collector or receiving device 2 for gathering harvested material from the ground can be suspended from the machine frame 3. This collector 2 can, for example, be designed as a pickup 3 with a rotating, driven spiked roller and a downstream conveying rotor 4 to collect harvested material such as hay, straw, or grass from the ground. The collector 2 can be guided over the ground by a sensing carriage 21 and pivoted relative to the machine frame 19.

[0037] The harvested crop collected from the ground by the collector or receiving device 2 is transferred to a bale former 22, which may have a bale forming chamber 7 in a manner known per se, in which incoming harvested crop can be formed into a bale by means of circulating belts 21 and / or circulating rollers. The bale former 22, including its bale forming chamber 7, may be mounted on the machine frame 3.

[0038] Downstream of the bale former 7 can be a wrapper 23, to which the bale former 22 can transfer a finished bale. The wrapper 23 can wrap the bale with a net, film, or other wrapping material before the wrapped bale is placed on the ground. For this purpose, the wrapper 23 can have a winding table 24 and winding elements 25 arranged around it, for example in the form of winding rollers, in a manner known per se (see Figure 1). Fig. 1.

[0039] How Fig. As shown in Figure 2, the aforementioned conveying rotor 4 can convey the crop flow 8 through a conveying channel 26 into the bale forming chamber 7, whereby cutting blades 6 of a cutting unit 5 can be arranged in the aforementioned conveying channel 26 for cutting the crop, cf. Fig. 2.

[0040] A sensor system 9 can be provided on the baler 1 in order to be able to detect various crop and operating parameters of the baler 1, in particular the moisture content of the crop flow taken up by the receiving device 2, the bale density of the crop bale forming in the bale forming chamber 7 and its bale diameter.

[0041] The sensor 9 can include a moisture sensor 10, which can be arranged, for example, in the area of ​​the aforementioned conveying channel 26 to measure the moisture content of the harvested crop flow 8. This at least one moisture sensor 10 can preferably include a conductivity and / or resistance meter to determine the moisture content by means of conductivity or resistance measurement. As explained above, however, capacitive moisture sensors, microwave moisture sensors, or near-infrared moisture meters can also be used.

[0042] The sensor system 9 also includes a bale density sensor 11, which can, for example, detect the belt tension of one or more belts 27 that encircle the forming bale in the bale forming chamber 7, cf. Fig. 2.

[0043] Preferably, the baler 1 can be designed as a variable baler or have a variable bale forming chamber 7, wherein the belts 27 can be tensioned and placed around the emerging bale of harvested crop in a variable size, cf. Fig. 2.

[0044] For belt tensioning, for example a belt tensioner 28 can be provided, which may have a pivotable tensioning arm 29 that carries a deflection pulley for the belts 27, cf. Fig. 2. The aforementioned tensioning arm 29 can be pivoted by a pressure medium actuator, for example in the form of a hydraulic cylinder 30, and thus the belt 27 can be tensioned, cf. Fig. 2.

[0045] The bale density sensor 11 mentioned above can be a pressure sensor that measures the pressure prevailing in the pressure medium actuator or the hydraulic cylinder 30, which is a measure of the bale density of the forming harvested crop bale.

[0046] To detect the bale diameter, the aforementioned sensor system 9 includes a diameter sensor 12, which can be arranged in the aforementioned bale forming chamber 7. Notwithstanding this, the aforementioned bale diameter sensor 12 can be designed to operate without contact, for example, in the form of an ultrasonic sensor, in order to determine or detect the distance between a mounting point and a bale surface and / or a belt 27 wrapping around the bale. This distance is a measure of the bale diameter, which can also be detected, for example, by means of one or more angle sensors and the resulting geometric relationships.

[0047] The sensor signals from sensor 9 are provided to a training module 13, possibly after preprocessing by a job computer or a signal filter that may be assigned to or associated with sensor 9.

[0048] The sensor 9 can also include a knife count sensor 33, which detects the number of cutting knives 6 in the cutting unit 5. Such a knife count sensor 33 can, for example, monitor the mounting positions for the cutting knives 6 and their occupancy, that is, detect whether a cutting knife is mounted in a respective mounting position or not.

[0049] Furthermore, the sensor 9 can also include a crop type sensor 34 to detect or determine the type of crop that is predominantly contained in or consists of the crop stream 8. The aforementioned crop type sensor 34 can monitor the crop stream 8 itself, or it can be attached to another agricultural machine, for example, a mower used to cut the crop or a combine harvester. In this case, the sensor 34 can transmit the acquired data remotely to the baler 1 or the evaluation unit 15 and / or the training module 13.

[0050] For example, the aforementioned crop type sensor 34 can include a camera device with an image evaluation device to optically detect the crop, wherein the image evaluation device can compare the detected images of the crop with stored reference images to determine the crop type.

[0051] For the training phase already described, the bales of harvested crop formed in various bale-forming processes can be measured for their bale weight using a weighing system 17. The weighing system 17 can, for example, be attached to the wrapper 23, perhaps in the form of a load cell in the area of ​​the wrapping table 24. However, it is also possible to use a separate weighing system 17 from the baler 1 to determine the bale weight during the training phase, for example, in the form of a separate weighing table on which the harvested crop bales can be placed.

[0052] The baler initially operates in a training mode, running through the various bale-forming processes, with the actual bale weight being measured by the weighing system 17. In later operating phases, the baler 1 can then operate without the weighing system 17 and still determine the bale weight.

[0053] However, it should be noted that data sets collected during work operations can also be stored in training memory 14 in order to further refine the evaluation of the data sets by AI or to further increase the data basis.

[0054] To determine the bale weights of the harvested crop bales formed by baler 1, a two-stage procedure is generally used. First, a method is used to determine the bale weight of harvested crop bales formed by an agricultural baler from a stream of harvested crop. In a training phase, the moisture content of the harvested crop stream 8, the bale density of the harvested crop bale formed in the baling chamber 7 of baler 1, and the bale diameter are recorded for several bale-forming processes using sensors 9. The weight of the formed harvested crop bales is then measured using a weighing system 17. Finally, the data sets from the training phase, containing the recorded moisture content, bale density, bale diameter, and measured bale weights, are stored in a training memory 14.The stored data sets are analyzed by an analysis device 15 and correlations between the measured bale weights and the recorded moisture content, bale density, and bale diameter are determined. In an operational phase following the training phase, the moisture content of the incoming crop flow, the bale density of the crop bale formed in the bale forming chamber 7 of the baler 1, and the bale diameter are recorded using the aforementioned sensor 9 for bale formation. The bale weight of the formed bale is then determined by the analysis device 15 based on the correlations previously determined in the training phase, from the moisture content, bale density, and bale diameter recorded in the operational phase, without a measurement signal from a weighing system.

[0055] Method according to the preceding claim, wherein the moisture content of the ingested crop flow, the bale density and the bale diameter are each determined in real time by the sensor 9 and the analysis device 15 evaluates the temporal course of the recorded moisture content, bale density and bale diameter during the bale forming process with regard to the bale weight measured by the weighing system and / or with regard to the bale weight to be determined without a signal from the weighing system.

[0056] Method according to the preceding claim, wherein the curves of the sensor signals of the sensor system 9, which indicate the moisture, bale density and bale diameter, are evaluated and / or integrated with respect to their slope and / or their curvature in order to determine a total moisture, total bale density and total bale diameter resulting over the duration of a bale forming process, wherein preferably, when determining the total moisture and the total bale density, the signal curves of the sensor system 9, which indicate the moisture and bale density, are each related to the curve of the signal of the bale diameter and thus to the current intermediate value of the bale diameter, in order to take into account in which diameter section which moisture and which bale density occurs.

[0057] Method according to one of the preceding claims, wherein in the training phase various bale formations are carried out with different numbers of cutting knives 6 of a cutting unit 5 for cutting the ingested crop flow and the respective number of cutting knives are stored in the training memory 14 for the data sets, wherein a relationship between the number of cutting knives and the bale weights measured by the weighing system 17 is determined from the data sets stored in the training memory 14, and wherein in the operating phase the bale weight is additionally determined on the basis of the number of cutting knives used in the operating phase.

[0058] A method according to one of the preceding claims, wherein in the training phase different types of crops are recorded for several bale formations and stored in the training memory 14 as data sets, wherein the evaluation unit determines a relationship between the types of crops of the recorded crop streams and the bale weights measured by the weighing system 17 from the data sets of the training memory 14, and wherein in the operating phase the crop type of the recorded crop stream is determined or entered and the evaluation unit 15 determines the bale weight in the operating phase based on the determined relationship between crop type and bale weight.

[0059] Method according to one of the preceding claims, wherein the evaluation device 15 subjects the data sets from the training memory 14 to a regression using a regression module and determines the relationship between moisture of the recorded crop flow, bale density and bale diameter on the one hand and bale weight on the other hand using regression.

[0060] Method according to one of the preceding claims, wherein the evaluation device 15 supplies the data sets of the training memory 14 to a neural network for determining the relationship between moisture of the recorded crop flow, bale density and bale diameter on the one hand and bale weight on the other.

[0061] Method according to one of the preceding claims, wherein the moisture content of the recorded crop flow 8 is detected by conductance and / or resistance measurement using a conductance and / or resistance sensor 10.

[0062] Method according to one of the preceding claims, wherein the bale density is detected in the training phase and in the operating phase by means of a force or pressure sensor 11, by which a belt tension of at least one belt in the bale forming chamber 7 of the baler is detected.

[0063] Method according to one of the preceding claims, wherein the bale diameter in the bale forming chamber 7 is detected by means of a non-contact distance and / or angle sensor, in particular an ultrasonic sensor, by which a distance between a belt encircling the harvested crop bale in the bale forming chamber 7 and a mounting fixed point is determined.

[0064] Method according to one of the preceding claims, wherein the type of crop of the crop stream 8 is automatically determined by means of a determination device, in particular by means of a camera system during mowing of the crop and / or a field record.

[0065] The aforementioned training module 13 includes a training memory 14 in which the transmitted sensor signals and / or information derived therefrom can be stored. The training module, including the training memory 14, can be part of the baler 1 or be designed separately, in which case a communication interface and / or a communication module for transmitting the sensor signals may be provided on the baler 1.

[0066] An evaluation unit 15 is provided for evaluating the data sets collected in the training memory 14, see below. Fig.1, which may include an electronic computing device, for example, in the form of a server or a computer with a processor, program memory, and main memory. In particular, the aforementioned evaluation device 15 may include an AI module 31 to determine the relationships between the moisture content of the harvested crop, bale density, and bale diameter on the one hand, and the measured bale weights on the other, as already explained. Such an AI module 31 may contain or access a neural network. Alternatively or additionally, the AI ​​module 31 may include a regression module 32 to evaluate the data sets by means of regression analysis. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] NL 20 10 169 C

[0003] EP 20 00 785 B1

[0003] EP 10 34 695 B1

[0003] EP 27 45 675 B1

[0003] EP 40 56 024 A1

[0003]

Claims

[1] A method for determining the bale weight of crop bales formed by an agricultural baler from a received crop stream, wherein - in a training phase by means of a sensor (9) for several bale forming processes each - the moisture of the crop stream (8), - the bale density of the crop bale formed in the bale forming chamber (7) of the baler (1), and - the bale diameter is recorded and the weight of the formed crop bales is measured by means of a weighing system (17), - the data records from the training phase including the recorded moisture levels, bale densities and bale diameters as well as the measured bale weights are stored in a training memory (14), - the stored data sets are analyzed by an analysis device (15) and relationships between the measured bale weights and the recorded moisture contents, bale densities and bale diameters are determined, and - in an operating phase following the training phase, by means of the said sensor system (9) for bale formation - the moisture content of the harvested crop stream, - the bale density of the crop bale formed in the bale forming chamber (7) of the baler (1), and - the bale diameter - are recorded and the bale weight of the formed bale is determined by the / an analysis device (15) on the basis of the relationships previously determined in the training phase from the moisture content, bale density and bale diameter recorded in the operating phase without a measuring signal from a weighing system. [2] Method according to the preceding claim, wherein the moisture content of the collected crop stream, the bale density and the bale diameter are each determined in real time by the sensor system (9) and the analysis device (15) evaluates the time course of the recorded moisture content, bale density and bale diameter during the bale forming process with regard to the bale weight measured by the weighing system and / or with regard to the bale weight to be determined without a signal from the weighing system when evaluating the data sets from the training phase and / or when determining the bale weight in the operating phase. [3] Method according to the preceding claim, wherein the courses of the sensor signals of the sensor system (9), which indicate the moisture content, the bale density and the bale diameter, are each evaluated and / or integrated with regard to their gradient and / or their curvature in order to determine a total moisture content, total bale density and total bale diameter resulting over the duration of a bale forming process, wherein preferably when determining the total moisture content and the total bale density, the signal courses of the sensor system (9), which indicate the moisture content and the bale density, are each related to the course of the signal of the bale diameter and thus to the current intermediate state of the bale diameter in order to take into account in which diameter section which moisture content and which bale density occurs. [4] Method according to one of the preceding claims, wherein in the training phase different bale formations are carried out with different numbers of cutting knives (6) of a cutting unit (5) for cutting the received crop stream and the respective number of cutting knives is stored in the training memory (14) for the data sets, wherein a relationship between the number of cutting knives and the bale weight measured by the weighing system (17) is determined from the data sets stored in the training memory (14), and wherein in the operating phase the bale weight is additionally determined based on the number of cutting knives used in the operating phase. [5] Method according to one of the preceding claims, wherein in the training phase different crop types are recorded for several bale formations and stored in the training memory (14) as data sets, wherein the evaluation device determines a relationship between the crop types of the recorded crop streams and the bale weights measured by the weighing system (17) from the data sets of the training memory (14), and wherein in the operating phase the crop type of the recorded crop stream is determined or entered and the bale weight is determined by the evaluation device (15) in the operating phase on the basis of the determined relationship between crop type and bale weight. [6] Method according to one of the preceding claims, wherein the evaluation device (15) subjects the data sets from the training memory (14) to a regression by means of a regression module and determines the relationship between moisture of the collected crop flow, bale density and bale diameter on the one hand and bale weight on the other hand with the aid of the regression. [7] Method according to one of the preceding claims, wherein the evaluation device (15) feeds the data sets of the training memory (14) to a neural network for determining the relationship between moisture of the collected crop flow, bale density and bale diameter on the one hand and bale weight on the other hand. [8] Method according to one of the preceding claims, wherein the moisture of the received crop stream (8) is detected by means of conductivity and / or resistance measurement by means of a conductivity and / or resistance sensor (10). [9] Method according to one of the preceding claims, wherein the bale density in the training phase and in the operating phase is detected by means of a force or pressure sensor (11) by which a belt tension of at least one belt in the bale forming chamber (7) of the baler is detected. [10] Method according to one of the preceding claims, wherein the bale diameter in the bale forming chamber (7) is detected by means of a contactless distance and / or angle sensor, in particular an ultrasonic sensor, by means of which a distance between a belt wrapping around the crop bale in the bale forming chamber (7) and a mounting fixed point is determined. [11] Method according to one of the preceding claims, wherein the crop type of the crop stream (8) is determined automatically by means of a determination device, in particular by means of a camera system when mowing the crop and / or a field record. [12] Device for determining the bale weight of crop bales formed by an agricultural baler (1) from a received crop stream (8), with a sensor system (9) for detecting - the moisture content of the collected crop stream (8), - the bale density of the crop bale in a bale forming chamber (7) of the baler (1), and - the bale diameter, - a training memory (14) for storing the data sets recorded by the sensor system (9), including the recorded moisture levels, bale densities and bale diameters for several bale formations carried out in a training phase, each linked to the bale weights measured by a weighing system (17), - an evaluation device (15) for evaluating the data sets stored in the training memory (14) and determining a relationship between the recorded moisture levels, the recorded bale densities and the recorded bale diameters on the one hand and the measured bale weights on the other hand, - wherein the / an evaluation device (15) is configured to determine the bale system from the moisture of the received crop stream (8) detected by the sensor system (9) in an operating phase, the bale density and the bale diameter of the crop bale formed in the operating phase, from the determined relationship between moisture, bale density and bale diameter on the one hand and the measured bale weight on the other hand, without the aid of a signal from the weighing system. [13] Device according to the preceding claim, wherein the evaluation device (15) comprises an AI module for determining said relationship between the bale weights measured in the training phase and the moisture contents, bale densities and bale diameters recorded in the training phase, wherein said AI module in particular comprises or accesses a neural network. [14] Agricultural baler with a receiving device (2) for receiving crop from the ground, a bale forming chamber (7) for forming the received crop into a crop bale, a sensor system (9) for detecting the moisture content of the crop stream received by the receiving device (2), the bale density of the crop bale in the bale forming chamber (7) and the bale diameter in the bale forming chamber (7), and an evaluation device (15) for determining the bale weight based on a relationship between moisture content, bale density and bale diameter on the one hand and bale weight on the other hand, which relationship can be provided to the evaluation device (15) by a device according to claims 1 to 10. [15] Baler according to the preceding claim, comprising a device according to claims 12 to 13.

Images