A method and system for optimizing spatial and temporal attribution of yield data of a harvester

Abstract

The application provides a harvester yield data space-time attribution optimization method and system, relates to the field of intelligent yield estimation technology of agricultural machinery, and comprises the following steps: acquiring weighing data, position data and state correlation data in the operation process of the harvester; identifying the operation state based on the state correlation data, and compensating the weighing data according to the operation state to obtain a weight net increment; establishing a time corresponding relationship between the weight net increment and the historical position data according to a delay alignment parameter; determining an operation coverage area according to the position data and an operation coverage parameter, and performing space deduplication to obtain an effective harvesting area; determining the yield per mu data based on the time corresponding relationship, the weight net increment and the effective harvesting area, cutting the yield attribution area under the head switching condition, and generating a yield distribution map.

Description

Technical Field

[0001] This application belongs to the field of intelligent yield measurement technology for agricultural machinery, specifically involving a method and system for optimizing the spatiotemporal attribution of harvester yield measurement data. Background Technology

[0002] With the development of intelligent agricultural machinery and precision agriculture technology, combine harvesters typically need to acquire real-time data such as yield, location, and operating area during operation, and generate yield distribution maps based on this data to reflect yield differences in different areas of the field. Weighing-based yield measurement has become one of the important methods in harvester yield measurement systems because it can directly reflect changes in crop weight within the grain bin.

[0003] Existing harvester yield measurement systems typically collect grain bin weight signals using weighing sensors, obtain the harvester's operating position via satellite positioning or integrated navigation, and calculate the harvested area by combining information such as the harvester's cutting width and travel trajectory. During data processing, signal processing methods such as mean filtering, low-pass filtering, and Kalman filtering are commonly used to smooth the weighing data, and the yield per unit area or average yield per acre is calculated based on the trajectory data.

[0004] However, during the dynamic operation of the harvester, the weighing signal is easily affected by factors such as vehicle vibration, ground undulation, changes in grain bin load, and changes in crop flow status. Additionally, there is a conveying time involved in the crop's movement from the harvesting unit to the weighing area, which may cause a time discrepancy between position data and weight changes. Furthermore, when the harvester turns around at the edge of the field, overlaps paths, or undergoes short-term start-stop switching, issues such as duplicate area statistics and inconsistent yield attribution may occur. Summary of the Invention

[0005] This application provides a method and system for optimizing the spatiotemporal attribution of harvester yield measurement data to solve one of the aforementioned technical problems.

[0006] The technical solution adopted in this application is as follows: This application provides a method for optimizing the spatiotemporal attribution of harvester yield measurement data, applied to the controller of a harvester yield measurement system. The method includes: Acquire multi-source yield measurement data during the harvester operation process, including weighing data, location data, and state correlation data used to characterize the harvester's operating status; The operating status of the harvester is identified based on the state association data, and the weighing data is compensated according to the operating status to obtain the net weight increment. The delay alignment parameters are determined based on the multi-source production measurement data and / or preset operation parameters, and a time correspondence between the net weight increment and historical location data is established based on the delay alignment parameters. The operating coverage area is determined based on the location data and the harvester's operating coverage parameters, and spatial deduplication is performed on the operating coverage area to obtain the effective harvest area; Based on the time correspondence, the net weight increment, and the effective harvested area, the yield per mu data is determined. When the operation status meets the field switching conditions, the yield distribution area corresponding to the yield per mu data is segmented to generate a yield distribution map.

[0007] According to one embodiment of this application, acquiring multi-source yield measurement data during harvester operation includes: The weighing data, the location data, and the state-related data are obtained according to a uniform sampling period, or time markers are added to the weighing data, the location data, and the state-related data collected at different sampling frequencies. Based on the time identifier, a sampling correspondence is established between the weighing data, the location data, and the status association data, and the operation status identification, the compensation processing, and the establishment of the time correspondence are performed based on the sampling correspondence.

[0008] According to one embodiment of this application, determining the delay alignment parameters based on the multi-source yield measurement data and / or preset operating parameters includes: Based on state-related data used to characterize crop flow rate, crop type parameters, crop moisture content parameters, and / or harvester conveyor structure parameters, determine the delay time corresponding to the crop entering the weighing area from the harvesting component; The delay time can be used as the delay alignment parameter, or the corresponding delay sampling number can be determined based on the delay time and the sampling period of the location data, and the delay sampling number can be used as the delay alignment parameter.

[0009] According to one embodiment of this application, identifying the operating status of the harvester based on the status association data includes: The movement status of the harvester is determined based on the location data, and the operating status of the harvesting component and the flow status of the harvesting process are determined based on the status association data. When the harvester is in a moving state, the harvesting component is in a working position, and the working material circulation state meets the harvesting circulation conditions, the harvester is identified as being in a harvesting operation state. If the harvesting component is detected to have switched from a working position to a non-working position, an exit delay process corresponding to the delay alignment parameter is initiated, and if the exit delay process has not ended, the harvester is identified as being in an exit transition state. If the exit delay process has not ended, and the harvesting component switches from a non-operational position to an operational position again and meets the harvesting flow conditions again, the harvester is identified as meeting the field switching conditions.

[0010] According to one embodiment of this application, the weighing data is compensated based on the operating status to obtain a net weight increment, including: When the operation status is a stationary state or a non-harvesting moving state, a first compensation parameter is determined based on the operation flow status data in the status association data, and the weighing data is compensated based on the first compensation parameter. When the operation status is harvesting operation status, a second compensation parameter is determined based on the vehicle posture data and grain bin load data in the status association data, and the weighing data is compensated based on the second compensation parameter; The net weight increment is determined based on the difference between the weighing data after compensation in adjacent sampling periods.

[0011] According to one embodiment of this application, a harvesting coverage area is determined based on the location data and the harvester's operating coverage parameters, and spatial deduplication is performed on the harvesting coverage area to obtain the effective harvested area, including: Based on the location data, the direction of travel of the harvester, and the operation coverage parameters, the operation coverage area corresponding to the harvesting component within the current sampling period is determined; The operation coverage area is mapped to a preset set of spatial units, and the target spatial units that are not included in the harvest area are determined according to the operation marking status of each spatial unit in the preset set of spatial units. The effective harvest area is determined based on the area corresponding to the target spatial unit, and the operation marker status of the target spatial unit is updated.

[0012] According to one embodiment of this application, establishing a time correspondence between the net weight increment and historical location data based on the delay alignment parameter includes: Write the location data, effective harvested area and time identifier corresponding to each sampling period into the cache queue in the sampling order; Based on the delay alignment parameter, read the historical location data and historical effective harvest area corresponding to the target historical period from the cache queue; The net weight increment corresponding to the current sampling period is associated with the historical location data and historical effective harvest area corresponding to the target historical period to form the time correspondence.

[0013] According to one embodiment of this application, when the operation status meets the field switching conditions, the yield region corresponding to the yield data per mu is segmented, including: When the harvester enters the exit transition state from the harvesting operation state, the net weight increase generated during the exit delay process corresponding to the exit transition state is attributed to the operation area before entering the exit transition state. If the exit delay process has not ended and the harvester meets the harvesting operation conditions again, the moment when the harvester meets the harvesting operation conditions again is taken as the re-entry start time, and the re-entry delay process corresponding to the delay alignment parameter is started. Before the re-entry delay process reaches the delay time or delay sampling number corresponding to the delay alignment parameter, the net weight increment generated during the re-entry delay process is temporarily stored in the transition data area or marked as field transition production data, and the net weight increment is restricted from being directly attributed to the work area after re-entry. After the re-entry delay process reaches the delay time or delay sampling number corresponding to the delay alignment parameter, the location data and historical effective harvest area corresponding to the target historical period are read from the cache queue according to the delay alignment parameter. The net weight increment generated after the re-entry delay process ends is associated with the location data and historical effective harvest area corresponding to the target historical period to determine the yield per acre data corresponding to the operating area after re-entry.

[0014] According to one embodiment of this application, the location data includes at least one of satellite positioning data, integrated navigation data, or location data determined based on the work trajectory; The status-related data includes at least two of the following: harvesting component height data, conveying component rotation speed data, conveying component load data, vehicle body tilt angle data, and vehicle speed data. The field switching conditions include combinations of at least two of the following: switching of harvesting components up and down, change of operating direction, change of moving speed, and change of crop flow status.

[0015] This application embodiment also provides a harvester yield measurement system, including a controller and a multi-source sensor communicatively connected to the controller. The multi-source sensor is used to collect multi-source yield measurement data during the harvester operation process, and the controller is used to execute the harvester yield measurement data spatiotemporal attribution optimization method.

[0016] Due to the adoption of the above technical solution, the beneficial effects achieved by this application are as follows: This application acquires weighing data, location data, and status-related data, and identifies the harvester's operating status based on the status-related data. This enables weighing compensation, delay alignment, area calculation, and field attribution processing to be linked based on the operating status, avoiding inconsistencies in data attribution caused by the independent processing of various yield measurement data.

[0017] This application compensates for weighing data based on the operational status, enabling weighing data in different states such as stationary, moving, harvesting, and transition to exit to use appropriate compensation logic. This reduces the impact of vehicle vibration, workpiece flow, and load changes on the net weight increment, providing a stable weight basis for subsequent yield per acre calculations.

[0018] This application determines the delay alignment parameter based on multi-source yield measurement data and / or preset operating parameters, and establishes a time correspondence between the net weight increment and historical location data based on the delay alignment parameter, so that the net weight increment generated in the current sampling period can be associated with the historical location data corresponding to the actual harvest of the crop, thereby reducing the time misalignment between yield data and location data.

[0019] This application determines the operational coverage area based on location data and operational coverage parameters, and performs spatial deduplication processing on the operational coverage area to prevent duplicate coverage areas from being counted twice in the effective harvest area, thereby ensuring that the area basis of the yield per acre data corresponds to the actual harvest area.

[0020] This application determines yield per acre data based on time correspondence, net weight increment, and effective harvested area, so that the yield per acre data simultaneously considers the temporal attribution of weight changes and the spatial attribution of harvested area, thereby improving the data bias caused by yield measurement based solely on current location and trajectory area.

[0021] This application, by dividing the yield data per acre into regions when the field switching conditions are met, allows the yield data under conditions such as turning around at the field, switching between header lifting and lowering, and re-entering the operation before the transition is over to be attributed to the corresponding operation area according to the state transition relationship, thereby reducing the cross-regional attribution of yield data in the field area.

[0022] This application enables weighing data, location data, and status-related data to maintain a traceable temporal relationship at different sampling frequencies by adding time stamps to multi-source yield measurement data and establishing sampling correspondence, thereby supporting subsequent status identification, weighing compensation, and time alignment processing.

[0023] This application records the location data, effective harvested area, and time identifier of each sampling period through a cache queue, and reads the data of the target historical period according to the delay alignment parameter. It can realize the correspondence between the net weight increment and the historical operation position with low computational complexity, providing a basis for the real-time operation of the controller. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating the multi-source data fusion optimization method provided in the embodiments of this application. Figure 2 This application provides a data processing pipeline diagram for a multi-source data fusion optimization method according to its embodiments. Figure 3 A flowchart of the area deduplication algorithm provided in the embodiments of this application; Figure 4 A cascaded filtering framework diagram provided for embodiments of this application; Figure 5 The state machine and output attribution logic diagram of the field optimization algorithm provided in the embodiments of this application are shown. Detailed Implementation

[0025] The technical solution of this application will be further described below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are only used to illustrate this application and are not intended to limit the scope of protection of this application.

[0026] In the description of the embodiments of this application, the terms "first," "second," etc., are only used to distinguish different objects and should not be construed as indicating or implying relative importance, nor should they be construed as limiting the quantity or processing order. The term "multiple" can mean two or more. The term "and / or" is used to describe the association relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can mean that only A exists, both A and B exist, or only B exists.

[0027] In this embodiment, the harvester can be a combine harvester, a tracked harvester, or an agricultural harvesting machine with weighing and yield measurement functions. The crop can include grains, cereals, or other materials harvested by the harvester and transported through a conveyor to the weighing area. The weighing data can include the raw voltage signal output by the weighing sensor array, the converted weight value, the filtered weight estimate, or the compensated weight data. The location data can include GNSS positioning data, BeiDou positioning data, RTK positioning data, integrated navigation data, or location data determined based on the operational trajectory.

[0028] Status-related data refers to data used to characterize the operating status of a harvester, and may include at least one of the following: harvester height data, conveyor rotation speed data, conveyor load data, vehicle tilt angle data, and vehicle speed data. For example, harvester height data may be header height data, conveyor rotation speed data may include grain elevator rotation speed data and waste elevator rotation speed data, and vehicle tilt angle data may include pitch angle data and roll angle data.

[0029] Operational coverage parameters refer to parameters used to determine the ground area covered by the harvester per unit time, and may include at least one of the following: cutting width, working width of the harvesting unit, direction of travel, operating speed, machine attitude correction parameters, and field boundary parameters. Effective harvested area refers to the net increase in harvested area used in yield calculations after removing duplicate coverage from the current operational coverage area. Delay alignment parameters are parameters used to describe the delay required for the crop to enter the weighing area from the harvesting unit; these can be a dynamic delay time or the number of delay samples corresponding to that dynamic delay time.

[0030] The time correspondence refers to the relationship between the net weight increment in the current sampling period and the location data and / or historical effective harvested area in historical sampling periods. Field-end switching conditions refer to the conditions that must be met when the harvester switches operating states in the field-end area; these conditions can be determined by at least two of the following: header lifting status, vehicle speed change, travel direction change, elevator speed change, and position trajectory turning. Yield attribution area refers to the operating area to which the net weight increment or yield per acre data should be assigned on the yield distribution map; this can be an operating row, trajectory segment, grid area, or other spatial unit within the field.

[0031] The exit delay process refers to the time window or sampling window initiated by the controller according to the delay alignment parameters after the harvesting unit switches from the working position to the non-working position; this process covers the transit time of crops from the harvesting unit through the conveyor channel to the weighing area. The exit transition state refers to the transitional control state identified by the controller to continue processing the weight assignment of crops in transit before the exit delay process has reached the delay time or delay sampling number corresponding to the delay alignment parameters; the exit transition state does not mean that the harvesting unit must remain in the non-working position throughout the entire state. Re-entering the delay process refers to the delay window initiated by the controller starting from the moment the harvesting unit switches back from the non-working position to the working position and meets the harvesting operation conditions again before the exit delay process has ended.

[0032] This application provides a method and system for optimizing the spatiotemporal attribution of harvester yield measurement data. The method is applied to the controller of a harvester weighing and yield measurement system. The controller acquires weighing data, location data, and status-related data during harvester operation through multi-source sensors, and drives weighing compensation, delay alignment, area deweighting, field yield attribution segmentation, and yield distribution map generation based on the operation status identification results.

[0033] Figure 1 The flowchart of the multi-source data fusion optimization method provided in the embodiments of this application is shown. Figure 2 The diagram illustrates the data processing pipeline of the multi-source data fusion optimization method provided in this application embodiment. Figure 3 A flowchart of the area deduplication algorithm provided in an embodiment of this application is shown. Figure 4 A cascaded filtering framework diagram provided in an embodiment of this application is shown. Figure 5 The state machine and output attribution logic diagram of the field optimization algorithm provided in this application embodiment are shown.

[0034] like Figure 1 and Figure 2 As shown, the harvester yield measurement system may include a controller, an intelligent display terminal, and multi-source sensors that are communicatively connected to the controller. The multi-source sensors may include at least one of the following: a weighing sensor group, a GNSS receiver, a header height sensor, a grain elevator speed sensor, a waste elevator speed sensor, an tilt sensor, and a vehicle speed sensor.

[0035] The load cell array is used to collect weighing signals from the grain silo or weighing support structure. The GNSS receiver outputs vehicle speed signals and a sequence of latitude and longitude trajectory points. The header height sensor outputs the current header height signal. The grain elevator speed sensor outputs the real-time speed signal of the grain elevator. The waste elevator speed sensor outputs the real-time speed signal of the waste elevator. The tilt sensor outputs the current pitch and roll angle signals of the harvester. The vehicle speed sensor can be set independently or the vehicle speed signal can be output from the GNSS receiver.

[0036] The controller can be an on-board controller, yield control module, electronic control unit, or processing unit with data acquisition and calculation capabilities within a harvester weighing and yield measurement system. The controller performs multi-source data acquisition, status recognition, weighing compensation, dynamic delay alignment, area deweighting, yield per acre calculation, cascade filtering, dynamic display, field yield cutting, yield distribution map generation, and data synchronization. The intelligent display terminal receives and displays the current weight, real-time yield per acre, cumulative yield, operating area, and yield distribution map output by the controller.

[0037] In one possible implementation, the controller deploys the method of this application embodiment in the form of a software program. Data collected by multiple sensors can be transmitted to the controller via CAN bus, serial port, Ethernet, analog acquisition circuit, or digital acquisition circuit. The controller can read data output by different sensors according to a uniform sampling period, or add time stamps to the data of each sensor at different sampling frequencies, and then establish the sampling correspondence between weighing data, position data, and status-related data based on the time stamps.

[0038] like Figure 1 As shown, the method provided in this application embodiment may include: S100, multi-source yield measurement data acquisition; S200, identifying the harvester's operating status based on the multi-source yield measurement data, and performing compensation processing on the weighing data according to the operating status to obtain the net weight increment; S300, determining the delay alignment parameter based on the multi-source yield measurement data and / or preset operating parameters, and establishing a time correspondence between the net weight increment and historical location data based on the delay alignment parameter; S400, determining the operating coverage area based on the location data and the harvester's operating coverage parameters, and performing spatial deduplication processing on the operating coverage area to obtain the effective harvest area; S500, determining the yield per mu data based on the time correspondence, the net weight increment, and the effective harvest area, and, when the operating status meets the field switching conditions, cutting the yield distribution area corresponding to the yield per mu data to generate a yield distribution map.

[0039] S100, acquire multi-source yield measurement data during the harvester operation process, the multi-source yield measurement data includes weighing data, location data and state association data used to characterize the harvester operation status.

[0040] Specifically, the controller acquires weighing data, location data, and status-related data in real time. Weighing data characterizes changes in crop weight within the harvester's grain bin and can be output by the weighing sensor array. Location data characterizes the harvester's operating position in the field and may include latitude and longitude coordinates, trajectory points, direction of travel, and speed. Status-related data characterizes the harvester's current operating status and may include header height, grain elevator speed, waste elevator speed, vehicle pitch angle, vehicle roll angle, vehicle speed, load on conveyor components, and other data related to the operating status.

[0041] In one possible implementation, the controller acquires data from each sensor according to a uniform sampling period, and treats the weighing data, position data, cutter height data, elevator speed data, and tilt angle data within the same sampling period as a set of multi-source production measurement data. In another possible implementation, different sensors have different data sampling frequencies. The controller adds a timestamp to each data point and maps data from different sources to the same processing period based on the timestamps. For data with inconsistent sampling times, nearest neighbor matching, preserving the previous valid value, linear interpolation, or resampling based on the sampling period can be used to establish the sampling correspondence.

[0042] Through the above processing, the controller can generate a multi-source yield measurement data sequence with time stamps. The weighing data in this multi-source yield measurement data sequence is used for subsequent weighing compensation and net weight increment calculation, the location data is used for subsequent operation coverage area calculation and historical location alignment, and the status association data is used for subsequent operation status identification, compensation strategy selection, and delay alignment parameter determination.

[0043] S200, based on the state association data, identify the operating state of the harvester, and perform compensation processing on the weighing data according to the operating state to obtain the net weight increment.

[0044] Specifically, the controller constructs a harvester operation status identification model based on state correlation data from multi-source yield measurement data. This operation status identification model can use preset multi-threshold joint judgment logic to identify the instantaneous operating condition of the harvester as one of the following: stationary state, normal harvesting state, exit transition state, normal exit state, and advance state before exiting. These states can be denoted as S0, S1, S2, S3, and S4, respectively.

[0045] S0 represents a stationary state. This state can indicate that the harvester speed is below a preset speed threshold, the header is in the raised position, and the elevator rotation speed is below a preset rotation speed threshold. In a supplementary implementation, if the harvester speed is below the preset speed threshold, but the engine, grain elevator, and waste elevator are still operating, the controller can identify this condition as a stationary component operating state and treat it as a compensation branch for a stationary or non-harvesting operation state. Therefore, the state identification conditions for a stationary state can be consistent with the compensation scenario for a stationary state.

[0046] S1 represents the normal harvesting state. This state indicates that the vehicle speed is higher than the preset speed threshold, the header height is lowered to the working position, and the elevator rotation speed is higher than the normal harvesting threshold. At this time, the controller considers the harvester to be performing effective harvesting operations and incorporates the weighing data, location data, and effective harvested area data into the normal yield measurement calculation process.

[0047] S2 is the exit transition state, also known as the delayed exit state. This state does not mean the harvesting unit must remain in a non-operating position throughout the entire state. Rather, it means that after the controller detects the harvesting unit switching from an operating position to a non-operating position, it initiates an exit delay process corresponding to the delay alignment parameter, and this exit delay process has not yet reached the dynamic delay time T or the corresponding delay sampling number. This state indicates that the header has triggered exit, but there are still crops en route at the elevator, conveyor channel, or grain bin entrance. The controller still needs to perform weighing compensation, delay alignment, and yield attribution processing so that the net weight increase generated by the crops en route can be attributed to the operating area before entering the exit transition state.

[0048] S3 is the normal exit state. This state indicates that the header has been raised and the dynamic delay time T has elapsed. The controller determines that the crops in transit for this work row have been processed, and the current operation is completely finished. After entering S3, the controller can either stop assigning yield to the current work row or wait for the header to lower again before entering a new normal harvest state.

[0049] S4 is the "advance before retreating" state. This state indicates that before the exit delay process corresponding to the S2 exit transition state has ended, the header switches back from a non-operating position to an operating position, and parameters such as machine speed, elevator speed, and workpiece flow status once again meet the judgment conditions for the normal harvesting state of S1. In other words, "advance before retreating" means that the harvester has already re-entered the harvesting conditions for the next operating row before the exit delay process of the previous operating row has ended. This state usually occurs in scenarios where the headland space is narrow, the operator makes a quick turn, or the operating row switching is rapid.

[0050] In one possible implementation, the field-end turning situation can be viewed as a state sequence S1→S2→S3→S1, or it can form a state sequence S1→S2→S4→S3→S1 in the case of a rapid turn. The controller can determine whether the harvester has entered the field-end area or whether the field-end switching conditions are met based on changes in header height, grain elevator speed, vehicle speed, direction of travel, and working position.

[0051] After identifying the operational status, the controller selects the corresponding weighing compensation strategy based on that status. The compensation process can target either the raw weighing signal output by the weighing sensor array or the weighing data after preliminary filtering. The output of the compensation process is the compensated weight data. The controller determines the net weight increment for the current sampling period based on the difference between the compensated weight data from adjacent sampling periods. This net weight increment is used to establish a time correspondence with historical location data and for calculating yield per acre and determining the yield's regional classification.

[0052] For segmented compensation in static state S0 or when stationary components are in operation, when the harvester is stationary or at low speed but the working components are still running, the main sources of interference include mechanical vibrations generated by the engine, grain elevator, and waste elevator. In this case, the controller can perform segmented compensation based on the rotational speeds of the grain elevator and waste elevator. Specifically, compensation coefficients corresponding to different speed ranges can be pre-calibrated through bench tests, forming a two-dimensional lookup table. The controller determines the corresponding compensation coefficient in the two-dimensional lookup table based on the measured rotational speeds of the grain elevator and waste elevator, and corrects the weight data based on this compensation coefficient to obtain the compensated weight data in the static state.

[0053] For segmented compensation during movement, when the harvester is in normal harvesting state S1 or other moving states, the main sources of interference include vehicle sway caused by field terrain, vehicle tilt, and weighing deviation caused by changes in grain distribution in the grain bin with the loading amount. In this case, the controller can perform joint segmented compensation based on the current cumulative weight of grain in the grain bin, the vehicle pitch angle, and the vehicle roll angle. Specifically, a multidimensional compensation coefficient matrix can be calibrated through orthogonal tests with different loading amounts and tilt angles. The controller then searches for the corresponding compensation amount in the multidimensional compensation coefficient matrix based on the measured weight, pitch angle, and roll angle, or substitutes it into a pre-fitted multivariate compensation function to calculate the compensated weight data in real time.

[0054] In one possible implementation, the controller can also set dedicated compensation strategies for the S2 exit transition state and the S4 pre-exit / entry state. In the S2 exit transition state, although the harvesting component has triggered the exit action, there are still crops in transit in the conveyor channel. The controller can continue to compensate for the weight data based on the conveyor component's rotation speed and the vehicle's posture. In the S4 pre-exit / entry state, the controller maintains the exit delay process corresponding to the previous work row and the re-entry delay process corresponding to the next work row, and manages the net weight increment within different delay processes in groups, avoiding directly using the net weight increment within the previous exit delay process as the basis for the yield per acre data of the next work row.

[0055] In one possible implementation, if the weighing data for a certain sampling period is invalid, or the status association data is insufficient to identify the operation status, the controller can use the previous valid operation status as the temporary status for the current period, or mark the current period as a period to be confirmed and suspend the update of the yield per acre data for that period. After obtaining valid status association data in subsequent sampling periods, the controller can re-determine the operation status and net weight increment attribution for the corresponding period based on the time stamp.

[0056] S300, determine the delay alignment parameter based on the multi-source production measurement data and / or preset operation parameters, and establish the time correspondence between the net weight increment and historical position data based on the delay alignment parameter.

[0057] Specifically, after crops are cut from the header, they need to be transported to the grain bin or weighing area via an elevator or other conveyor. Therefore, the net weight increment generated by the weighing data in the current sampling period usually corresponds to crops harvested at a previous historical moment or location. To avoid directly pairing the current net weight increment with the current location data, the controller introduces a dynamic delay time T and establishes a FIFO buffer alignment mechanism based on the dynamic delay time T.

[0058] The dynamic delay time T is not a fixed constant, but is dynamically determined based on at least one of the following: crop variety, real-time rotation speed of the grain elevator, grain moisture content, conveying structure parameters, and operating speed. The crop variety can be preset by the user in the task description. Grain moisture content can be collected by a moisture sensor, input by the user, or retrieved from the operating parameter table. The real-time rotation speed of the grain elevator can be collected by a grain elevator rotation speed sensor. The controller can determine the dynamic delay time T through a preset mapping relationship, a lookup table, or a preset function.

[0059] In one possible implementation, the controller determines a basic delay range based on the crop variety, a conveying speed correction based on the real-time rotation speed of the grain elevator, and a material flowability correction based on the grain moisture content, thereby obtaining the dynamic delay time T corresponding to the current processing cycle. If the location data is output with a fixed sampling period, the controller can convert the dynamic delay time T into a number of delayed samples and use this number of delayed samples as a delay alignment parameter.

[0060] The controller maintains a FIFO buffer queue in memory with a duration corresponding to the dynamic delay time T. At a fixed frequency, the system pushes the acreage data calculated from the location trajectory and harvesting width, the corresponding GPS or other location coordinates, the effective harvested area, the start coordinates, the end coordinates, the work row identifier, and the timestamp into the tail of the FIFO buffer queue for each unit of time. The data in the FIFO buffer queue is arranged in order of sampling time, and data can be read from the head of the queue or the target historical position according to the dynamic delay time T.

[0061] At any given moment, the net weight increment calculated by the controller is not directly paired with the location coordinates of that moment. Instead, it is paired with historical acreage data, historical location data, and historical effective harvested area in the FIFO buffer queue, corresponding to the current moment minus the dynamic delay time T. This establishes a temporal correspondence between the net weight increment and historical location data, allowing the yield data to be attributed to the historical geographical location that generated that yield.

[0062] When the dynamic delay time T changes, the controller can adjust the read position of the target historical period in the FIFO buffer queue. For example, when the dynamic delay time T increases, the controller reads the position data and effective harvested area of ​​an earlier historical period; when the dynamic delay time T decreases, the controller reads the position data and effective harvested area of ​​a more recent historical period. If the target historical period corresponding to the dynamic delay time T does not exist, the controller can temporarily store the current net weight increment and establish the time correspondence after the buffer queue meets the read conditions.

[0063] After the harvester enters the S2 exit transition state, although the harvesting component has triggered the exit action from the working position to the non-working position, the controller continues to run the yield measurement algorithm according to the exit delay process until the dynamic delay time T or the corresponding delay sampling number is reached. During this exit delay process, crops en route in the elevator may still enter the grain bin and cause changes in the weighing data. The controller continues to match the net weight increment generated during this exit delay process with the historical position data of the end area of ​​the previous working row to ensure that crops en route in the conveyor channel are included in the total yield and belong to the last working path.

[0064] S400: Determine the operating coverage area based on the location data and the harvester's operating coverage parameters, and perform spatial deduplication processing on the operating coverage area to obtain the effective harvest area.

[0065] Specifically, the controller determines the ground area covered by the harvesting components within the current sampling period based on location data, the harvester's travel direction, and operational coverage parameters. Operational coverage parameters may include at least one of the following: cutting width, harvesting component working width, machine attitude correction parameters, operational direction parameters, and field boundary parameters.

[0066] like Figure 3 As shown, in one possible implementation, the controller establishes a Cartesian coordinate system covering the entire work area, using square grids of a preset size as the basic spatial unit. For example, the work area can be divided into multiple equally sized square grids, each 1m x 1m in the controller's memory. Each grid corresponds to a Boolean flag in the controller's memory, initially set to FALSE, indicating that the grid has not yet been included in the harvest area; once the grid is included in the harvest area, the corresponding flag is updated to TRUE.

[0067] Within each sampling period, the controller calculates the ground strip area covered by the header at the current moment based on the real-time position coordinates obtained from GNSS positioning and the harvester's cutting width. The controller then maps this ground strip area onto a raster map to determine the set of grid cells that are completely or partially covered by the ground strip area.

[0068] For each grid cell in the grid set, the controller checks its Boolean flag. If the flag is FALSE, the controller marks the grid as TRUE and adds its area to the current period's effective harvested area. If the flag is TRUE, it means the grid has already been included in the harvested area, and the controller skips the grid and does not include its area in the current period's effective harvested area.

[0069] For partially covered grids, the controller can employ at least two processing methods. In one method, when the coverage ratio reaches a preset threshold, the area of ​​the grid is included in the effective harvest area; when the coverage ratio does not reach the preset threshold, it is not included in the effective harvest area. In another method, the controller calculates the contribution area of ​​the grid based on the actual coverage ratio and includes this contribution area in the effective harvest area. These processing methods can be pre-configured according to the controller's computing power and yield measurement accuracy requirements.

[0070] The controller uses the accumulated effective harvested area for the current period as the acreage data, along with the corresponding timestamp, start coordinates, end coordinates, work row identifier, and location data, and writes it to the FIFO buffer queue. Therefore, the FIFO buffer queue does not simply store the trajectory area, but rather the effective harvested area after spatial deduplication. When calculating yield per acre subsequently, the net weight increment can be paired with the historical effective harvested area to prevent duplicate coverage areas from being counted twice.

[0071] In another possible implementation, spatial deduplication can also employ an area deduplication method based on geometric overlay analysis. Specifically, the controller can represent the operation coverage area of ​​each sampling period as a polygonal region and perform overlay analysis with the historically harvested operation areas to calculate and subtract the overlapping areas, thus obtaining the effective harvested area for the current period. This method can serve as an alternative to the rasterized area deduplication method.

[0072] S500: Based on the time correspondence, the net weight increment, and the effective harvested area, determine the yield per mu data, and when the operation status meets the field switching conditions, cut the yield to which the yield per mu data belongs and generate a yield distribution map.

[0073] Specifically, the controller, based on the time correspondence, assigns the net weight increment corresponding to the current sampling period to the historical position data corresponding to the target historical period in the FIFO buffer queue, and determines the yield per acre data corresponding to that historical position or historical region by combining it with the historical effective harvested area corresponding to the target historical period. The effective harvested area here can be the data obtained by spatial deduplication in the current period and written to the buffer queue, or it can be the historical effective harvested area read from the buffer queue according to the delay alignment parameter. For real-time calculation of yield per acre data, the historical effective harvested area aligned with the completion time of the net weight increment is preferably used as the area basis.

[0074] like Figure 5 As shown, in one possible implementation, when the historical effective harvested area is greater than a preset area threshold, the controller determines the yield per acre data based on the time-aligned net weight increment and the historical effective harvested area. When the historical effective harvested area is less than or equal to the preset area threshold, or when the historical effective harvested area is zero, the controller can pause the calculation of the yield per acre data for the current period, temporarily store the current net weight increment in the transition data area, or mark the period as an invalid area period to avoid abnormal fluctuations in yield per acre data due to excessively small or zero area. The net weight increment temporarily stored in the transition data area can be assigned to the corresponding historical location area according to the time identifier and operation status when the subsequent effective harvested area meets the calculation conditions.

[0075] The controller determines whether to execute production allocation zone cutting based on the field conditions. Field conditions can be determined by real-time analysis of changes in data from the header height sensor, grain elevator speed sensor, and vehicle speed. Field switching conditions can include combinations of at least two of the following: header raising, header lowering, change of operating direction, change of vehicle speed, change of elevator speed, and position trajectory turning.

[0076] In a standard field turnaround process, when the operator raises the header at the field edge, the controller detects that the harvesting unit has switched from the working position to the non-working position and identifies the state as transitioning from S1 (normal harvesting) to S2 (exit transition). At this time, the controller initiates an exit delay process and continues to run the yield measurement algorithm according to the dynamic delay time T or the number of delay samples to process the crops still in transit in the elevator. The net weight increase generated by these crops is attributed to the end area of ​​the current work row. After the exit delay process reaches the dynamic delay time T or the corresponding number of delay samples, the controller identifies the state as entering S3 (normal exit). If the header subsequently lowers again and meets the normal harvesting conditions again, the controller switches the harvester state back to the new S1 (normal harvesting) state and starts a new delay timer.

[0077] In the "no retreat, no advance" condition, if the field space is narrow or the operator quickly turns around, causing the harvester to enter the S2 exit transition state and the exit delay process has not yet reached the dynamic delay time T or the delay sampling quantity, the header will switch back from the non-operating position to the operating position and meet the S1 normal harvesting conditions again. The controller then recognizes the harvester as entering the S4 "no retreat, no advance" state. For this condition, the controller performs spatiotemporal allocation and cutting of the yield before and after the delay process.

[0078] Specifically, the controller records the previous exit delay process as window T1, and fully attributes the net weight increment generated within window T1 to the first work row until the delay time T1 or the number of delay samples corresponding to the first work row has completely ended. Even if the cutting table is lowered again before window T1 has ended, the controller will not attribute the net weight increment within window T1 to the second work row.

[0079] When the header lowers again and meets the operating conditions once more, the controller uses this moment as the re-entry start time, initiates the re-entry delay process corresponding to the delay alignment parameter, and records this re-entry delay process as the T2 window. Before the T2 window reaches the dynamic delay time T or the corresponding delay sampling number, even if the crop has started entering the grain bin and causing changes in weighing data, the controller may temporarily not calculate the real-time yield per acre of the second operating row, or temporarily store the net weight increment within the T2 window in the transition data area and mark it as the field transition yield data.

[0080] Once the T2 window reaches the dynamic delay time T or the corresponding number of delayed samples, the controller determines that the harvesting process for the second work row has been established. For the net weight increment generated after the T2 window ends, the controller reads the location data, historical effective harvested area, and work row identifier corresponding to the target historical period from the FIFO cache queue according to the delay alignment parameter, and associates the net weight increment with the location data and historical effective harvested area corresponding to the target historical period to determine the yield per acre data corresponding to the second work row. For the net weight increment temporarily stored within the T2 window, the controller can either merge it into the starting effective area of ​​the second work row according to the preset transitional attribution rules, or continue to mark it as transitional yield data at the field edge. Thus, the net weight increment within the T2 window will not be lost, nor will it directly contaminate the real-time yield per acre calculation of the first or second work row.

[0081] Yield distribution maps can be generated by the controller based on historical location data, effective harvested area, and corresponding yield per acre. The controller can map yield per acre data to grid cells, trajectory segments, polygonal regions, or work rows, and generate field yield distribution maps in the form of color, numerical values, grades, or equal-value areas. The generated yield distribution maps can be sent to a smart display terminal for real-time display, or stored in local storage and uploaded to a cloud platform for subsequent analysis or archiving after the operation is completed.

[0082] like Figure 4 As shown, cascaded filtering and dynamic display processing are performed on the weighing and display data. Specifically, considering the characteristics of the original weighing sensor signal containing multiple composite noises, the controller can execute a four-stage cascaded filtering framework to purify the original weighing signal step by step. This four-stage cascaded filtering framework can include outlier removal, mean filtering, low-pass filtering, and Kalman filtering.

[0083] The first level is outlier removal. The controller monitors the raw voltage or weight signals output by the weighing sensor in real time. If a sampled value exceeds a preset physical reasonable range, or if the jump amplitude between adjacent sampled points exceeds a preset sudden change threshold, the controller determines that the sampled value is an outlier. For outliers, the controller can either discard them directly and replace them with the previous valid value, or mark the outlier as invalid data and exclude it from subsequent compensation and yield calculation.

[0084] The second stage is mean filtering. The controller sets a sliding window of width N, and performs an arithmetic mean on N consecutive weighing data points within the window to smooth out high-frequency random noise introduced by factors such as electromagnetic interference and sampling disturbances. The sliding window width N can be preset according to the sampling frequency, the vibration characteristics of the harvester, and the computing power of the controller.

[0085] The third stage is low-pass filtering. The controller uses a low-pass filter with a configurable cutoff frequency to process the weighing data after mean filtering, in order to filter out the periodic mechanical vibration noise generated by the operation of working components such as the engine and elevator. The cutoff frequency of the low-pass filter can be set according to the engine speed, elevator speed, or a pre-calibrated vibration frequency band.

[0086] The fourth stage is the Kalman filter. The controller establishes the system state equation and observation equation with weight as the state variable, and uses the weighing data processed by the first three stages as the observation input. The Kalman filter, through prediction and update loops, fuses the system state model with noisy observations and recursively outputs an estimate of the current true weight. This estimate can be used as input for subsequent weighing compensation, net weight increment calculation, or dynamic display.

[0087] In an alternative implementation, some filters in the four-stage cascaded filtering framework can be replaced by other filtering or estimation algorithms. For example, unscented Kalman filtering, adaptive strong-tracking unscented Kalman filtering, or intelligent compensation algorithms based on neural networks can be used instead of ordinary Kalman filtering; the number of filtering stages can also be reduced or adjusted according to the computing power of the harvester hardware platform. All of these alternative methods can achieve fine processing of weighing signals without changing the main workflow of spatiotemporal attribution processing of weighing data.

[0088] A dynamic display algorithm is used to handle sudden changes in display on the intelligent display terminal. When a combine harvester performs rapid unloading or high-feeding harvesting, causing a sharp change in the weight inside the grain bin, the controller performs a threshold judgment on the rate of change of the weight estimate output after cascaded filtering. If the weight change within two adjacent display update cycles does not exceed the preset display threshold, the controller can send the updated weight value to the intelligent display terminal for direct display. If the weight change within two adjacent display update cycles exceeds the preset display threshold, the controller does not directly send the updated weight value to the display screen. Instead, it stretches the weight change on the time axis, smoothly increasing or decreasing it over multiple time intervals until the displayed value catches up with the actual measured value or weight estimate.

[0089] The dynamic display algorithm is only used for smoothing the display on the terminal and does not change the actual measured values ​​or weight estimates used internally by the controller for yield calculation and output attribution. In one implementation, the controller still uses the compensated and filtered weight data for net weight increment calculation, while the weight value displayed on the terminal can be updated gradually according to the dynamic display algorithm.

[0090] In one possible implementation, the controller can execute the above method according to the following data processing pipeline: First, the operation status identification module outputs S0, S1, S2, S3, or S4 status based on multi-source sensor data; second, the dynamic and static compensation module compensates for the weighing data according to the operation status and outputs the net weight increment; third, the yield location time alignment module reads historical location data and historical effective harvested area from the FIFO buffer queue according to the dynamic delay time T, and pairs the net weight increment with the historical location data; subsequently, the area deduplication module outputs the effective harvested area according to the grid marker status; finally, the field optimization module divides the yield attribution area according to the state transition relationship and outputs the yield distribution map.

[0091] It should be noted that the time alignment processing in S300 and the spatial deduplication processing in S400 can be implemented in different execution orders. In one implementation, the controller first calculates the effective harvested area for the current period and writes it into the FIFO buffer queue, then reads the historical effective harvested area and pairs it with the current net weight increment according to the dynamic delay time T. In another implementation, the controller can perform the enqueueing of the effective harvested area for the current period and the reading of historical data in parallel within the same sampling period, as long as it can ensure that the area data used for yield calculation and the net weight increment have the same temporal attribution relationship.

[0092] This application also provides a data synchronization implementation method. After generating the yield distribution map, the controller can upload at least one of the following to a cloud platform: weighing data, location data, status-related data, effective harvested area, yield per acre data, yield distribution map, operation time, field identifier, and operation row identifier. The cloud platform can be used to store operation records, display field yield distribution maps, or perform operation quality traceability. In offline operation, the controller can first store the relevant data in local storage and perform cloud synchronization after communication is restored.

[0093] This application also provides an alternative source of raw yield data. While the above embodiments use weighing data output from a weighing sensor array as the primary raw yield data, in other possible implementations, the raw yield data can also come from photoelectric yield sensors, impulse yield sensors, radiation yield sensors, or other sensors capable of characterizing crop flow or yield changes. In this case, the controller can use the yield-related data output from the aforementioned sensors as alternative inputs to weighing data or yield characterization data, and continue to perform state recognition, delay alignment, spatial deduplication, and field assignment processing.

[0094] This application also provides a device embodiment. The harvester yield measurement data spatiotemporal attribution optimization device may include a data acquisition module, a status recognition module, a weighing compensation module, a delay alignment module, an area de-duplication module, a yield per acre determination module, a field edge cutting module, a filtering display module, and an image generation module.

[0095] The data acquisition module acquires multi-source yield measurement data during harvester operation. The status recognition module identifies the harvester's operating status based on status-related data. The weighing compensation module compensates for the weighing data according to the operating status to obtain the net weight increment. The delay alignment module determines the delay alignment parameters based on multi-source yield measurement data and / or preset operating parameters, and establishes a time correspondence between the net weight increment and historical location data based on the delay alignment parameters. The area deduplication module determines the operating coverage area based on location data and operating coverage parameters, and performs spatial deduplication on the operating coverage area to obtain the effective harvested area. The yield per mu (unit of land area) determination module determines the yield per mu data based on the time correspondence, net weight increment, and effective harvested area. The field-end cutting module cuts the yield-assigned area corresponding to the yield per mu data when the operating status meets the field-end switching conditions. The filtering and display module performs cascaded filtering and dynamic display processing. The image generation module generates a yield distribution map and outputs the yield distribution map to a smart display terminal or cloud platform.

[0096] This application also provides a harvester yield measurement system. The harvester yield measurement system may include a controller, an intelligent display terminal, and multi-source sensors communicatively connected to the controller. The multi-source sensors are used to collect multi-source yield measurement data during harvester operation. The controller is used to execute the harvester yield measurement data spatiotemporal allocation optimization method described in any of the above embodiments. The intelligent display terminal is used to display the current weight, yield per acre, effective harvested area, operating status, and yield distribution map output by the controller.

[0097] This application also provides an electronic device, which may include a processor and a memory. The memory stores a computer program, and the processor executes the computer program stored in the memory to implement the harvester yield measurement data spatiotemporal attribution optimization method described in any of the above embodiments. The electronic device may be a controller for a harvester weighing and yield measurement system, or a data processing device communicatively connected to the harvester yield measurement system.

[0098] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the harvester yield measurement data spatiotemporal attribution optimization method described in any of the above embodiments.

[0099] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

[0100] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0101] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for optimizing the spatiotemporal attribution of harvester yield measurement data, characterized in that, A controller applied to a harvester yield measurement system, the method comprising: Acquire multi-source yield measurement data during the harvester operation process. The multi-source yield measurement data includes weighing data, position data, and state-related data used to characterize the harvester's operating status. The state-related data includes at least two of the following: harvesting component height data, conveying component rotation speed data, conveying component load data, vehicle tilt angle data, and vehicle speed data. The harvester's operating status is identified based on the state association data. The operating status includes one of the following: stationary state, normal harvesting state, exit transition state, normal exit state, and pre-exit state. The pre-exit state indicates that the harvester has re-entered the harvesting of the next operation row before the exit delay process of the previous operation row has ended. The weighing data is compensated according to the operating status to obtain the net weight increment, including: When the operation status is static, a first compensation parameter is determined based on the operation flow status data in the status association data, and the weighing data is compensated based on the first compensation parameter. When the operation status is normal harvesting status, a second compensation parameter is determined based on the vehicle posture data and grain bin load data in the status association data, and the weighing data is compensated based on the second compensation parameter; The net weight increment is determined based on the difference between the weighing data after compensation in adjacent sampling periods; The delay alignment parameter is determined based on the multi-source yield measurement data and / or preset operation parameters, and a time correspondence between the net weight increment and historical location data is established based on the delay alignment parameter. The delay alignment parameter is the delay time corresponding to the crop entering the weighing area from the harvesting component, or the corresponding delay sampling quantity is determined based on the delay time and the sampling period of the location data. The operating coverage area is determined based on the location data and the harvester's operating coverage parameters, and spatial deduplication is performed on the operating coverage area to obtain the effective harvest area; Based on the time correspondence, the net weight increment, and the effective harvested area, the yield per mu data is determined. When the operation status meets the field switching conditions, the yield to which the yield per mu data belongs is segmented to generate a yield distribution map. The process of identifying the operating status of the harvester based on the status association data includes: If the harvesting unit is detected to have switched from a working position to a non-working position, an exit delay process corresponding to the delay alignment parameter is initiated, and if the exit delay process has not ended, the harvester is identified as being in an exit transition state.

2. The method according to claim 1, characterized in that, Acquire multi-source yield measurement data during harvester operations, including: The weighing data, the location data, and the state-related data are obtained according to a uniform sampling period, or time markers are added to the weighing data, the location data, and the state-related data collected at different sampling frequencies. Based on the time identifier, a sampling correspondence is established between the weighing data, the location data, and the status association data, and the operation status identification, the compensation processing, and the establishment of the time correspondence are performed based on the sampling correspondence.

3. The method according to claim 1, characterized in that, Determining delay alignment parameters based on the multi-source yield measurement data and / or preset operating parameters includes: Based on state-related data used to characterize crop flow rate, crop type parameters, crop moisture content parameters, and / or harvester conveyor structure parameters, determine the delay time corresponding to the crop entering the weighing area from the harvesting component.

4. The method according to claim 1, characterized in that, Identifying the operating status of the harvester based on the aforementioned status association data includes: The movement status of the harvester is determined based on the location data, and the operating status of the harvesting component and the flow status of the harvesting process are determined based on the status association data. When the harvester is in a moving state, the harvesting component is in a working position, and the working material circulation state meets the harvesting circulation conditions, the harvester is identified as being in a normal harvesting state. If the exit delay process has not ended, and the harvesting component switches from a non-operational position to an operational position again and meets the harvesting flow conditions again, the harvester is identified as meeting the field switching conditions.

5. The method according to claim 1, characterized in that, The operating coverage area is determined based on the location data and the harvester's operating coverage parameters, and spatial deduplication is performed on the operating coverage area to obtain the effective harvest area, including: Based on the location data, the direction of travel of the harvester, and the operation coverage parameters, the operation coverage area corresponding to the harvesting component within the current sampling period is determined; The operation coverage area is mapped to a preset set of spatial units, and the target spatial units that are not included in the harvest area are determined according to the operation marking status of each spatial unit in the preset set of spatial units. The effective harvest area is determined based on the area corresponding to the target spatial unit, and the operation marker status of the target spatial unit is updated.

6. The method according to claim 1, characterized in that, Establishing a time correspondence between the net weight increment and historical location data based on the aforementioned delay alignment parameters includes: Write the location data, effective harvested area and time identifier corresponding to each sampling period into the cache queue in the sampling order; Based on the delay alignment parameter, read the historical location data and historical effective harvest area corresponding to the target historical period from the cache queue; The net weight increment corresponding to the current sampling period is associated with the historical location data and historical effective harvest area corresponding to the target historical period to form the time correspondence.

7. The method according to claim 1, characterized in that, When the operation status meets the field switching conditions, the yield region corresponding to the yield data per mu is segmented, including: When the harvester enters the exit transition state from the normal harvesting state, the net weight increase generated during the exit delay process corresponding to the exit transition state is attributed to the working area before entering the exit transition state. If the exit delay process has not ended and the harvester meets the harvesting operation conditions again, the moment when the harvester meets the harvesting operation conditions again is taken as the re-entry start time, and the re-entry delay process corresponding to the delay alignment parameter is started. Before the re-entry delay process reaches the delay time or delay sampling number corresponding to the delay alignment parameter, the net weight increment generated during the re-entry delay process is temporarily stored in the transition data area or marked as field transition production data, and the net weight increment is restricted from being directly attributed to the work area after re-entry. After the re-entry delay process reaches the delay time or delay sampling number corresponding to the delay alignment parameter, the location data and historical effective harvest area corresponding to the target historical period are read from the cache queue according to the delay alignment parameter. The net weight increment generated after the re-entry delay process ends is associated with the location data and historical effective harvest area corresponding to the target historical period to determine the yield per acre data corresponding to the operating area after re-entry.

8. The method according to claim 1, characterized in that, The location data includes at least one of satellite positioning data, integrated navigation data, or location data determined based on the operation trajectory. The field switching conditions include combinations of at least two of the following: switching of harvesting components up and down, changes in operating direction, changes in moving speed, and changes in the state of crop flow.

9. A harvester yield measurement system, characterized in that, The system includes a controller and a multi-source sensor that is communicatively connected to the controller. The multi-source sensor is used to collect multi-source yield measurement data during the operation of the harvester. The controller is used to execute the spatiotemporal allocation optimization method for harvester yield measurement data as described in any one of claims 1 to 8.

Images