An agricultural machine operation quality automatic evaluation method, device and electronic equipment

Abstract

The application provides a kind of agricultural machinery operation quality automatic evaluation method, device and electronic equipment, it is related to agricultural machinery technical field, for solving because trajectory buffer area and actual plot profile deviation is larger, lack of work identification associated with track density characteristics, resulting in lower accuracy of operation quality evaluation defects, comprising: obtaining the preset operation width of agricultural machinery and operation track data, and according to the preset operation width, the first preset density clustering processing is carried out to operation track data, obtains target clustering cluster and initial noise point set;The target clustering cluster is carried out to the boundary extraction of concave hull, and the concave hull polygon boundary corresponding to target operation plot is obtained;Determine the missed operation area according to the concave hull polygon boundary and initial noise point set;According to operation track data, preset operation width, preset overlap area constraint condition, determine excess overlap area;According to missed operation area, excess overlap area and concave hull polygon boundary determine comprehensive quality evaluation result.

Description

Technical Field

[0001] This application relates to the field of agricultural machinery technology, and in particular to an automatic evaluation method, device and electronic equipment for agricultural machinery operation quality. Background Technology

[0002] The evaluation of agricultural machinery operation quality is an important part of precision agriculture management. By quantitatively assessing the actual coverage of agricultural machinery in the field, data support can be provided for operation scheduling, agricultural input accounting, and refined farmland management.

[0003] Currently, trajectory buffer spatial analysis is commonly used for evaluating agricultural machinery operation quality. This method generates a fixed-width strip buffer along the machinery's operation trajectory, using the union of these buffers as an approximation of the already completed area. Spatial operations are then performed on the buffers to calculate the missed and overlapping areas, thus obtaining evaluation indicators such as the missed operation rate and overlap rate. However, this method suffers from significant discrepancies between the buffer union and the actual plot outline when the plot boundaries are irregular or the machinery's operation path is non-standardized. Furthermore, the identification of missed operation areas relies entirely on the detection of geometric gaps within the buffers, lacking effective correlation with the density distribution characteristics of the operation trajectory itself. This leads to non-operational areas, such as field bends, being easily misjudged as missed operations, while the true missed operation areas remain difficult to locate accurately. Summary of the Invention

[0004] This application provides an automatic evaluation method, device, and electronic equipment for agricultural machinery operation quality, which solves the defects in the prior art where the trajectory buffer deviates significantly from the actual plot outline and the lack of correlation between missed operation identification and trajectory density features leads to low accuracy in operation quality evaluation. It realizes the automatic extraction of plot boundaries from operation trajectory data, the location of missed operation areas using noise point mapping, and the automatic completion of comprehensive quality evaluation.

[0005] In a first aspect, this application provides an automatic evaluation method for the quality of agricultural machinery operations, comprising the following steps: The preset operating width and operating trajectory data of agricultural machinery are obtained, and the operating trajectory data are subjected to the first preset density clustering processing according to the preset operating width to obtain the target cluster and the initial noise point set. Based on a preset polygon extraction algorithm, the concave hull boundary of the target cluster is extracted to obtain the concave polygon boundary corresponding to the target work plot. The missed work area is determined based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set; Based on the operation trajectory data, the preset operation width, and the preset overlap area constraints, the excessive overlap area is determined; The comprehensive quality evaluation results for the target work site are determined based on the missed work areas, the excessive overlap areas, and the concave polygon boundaries.

[0006] In one possible implementation, the work trajectory data is subjected to a first preset density clustering process based on a preset work width to obtain a target cluster and an initial set of noise points, including: The search neighborhood radius parameter corresponding to the first preset density clustering process is determined based on the preset operation width, and the minimum number of neighborhood samples corresponding to the first preset density clustering process is determined based on the preset expected point density. Spatial connectivity detection is performed on the operation trajectory data based on the search neighborhood radius parameter and the minimum number of samples in the neighborhood parameter; Trajectory data points that satisfy the density connectivity characteristic are aggregated into target clusters; Trajectory data points that do not meet the density connectivity feature are classified into the initial noise point set.

[0007] In another possible implementation, the concave hull boundary of the target cluster is extracted based on a preset polygon extraction algorithm to obtain the concave polygon boundary corresponding to the target work area, including: Based on the preset working width, determine the concave hull control parameters corresponding to the preset polygon extraction algorithm; The target cluster is input into a preset polygon extraction algorithm, and the initial concave hull polygon outline is constructed by combining the concave hull control parameters. Detect the empty areas inside the initial concave polygon outline, retain the empty areas with an area greater than a preset area threshold as non-operational areas within the plot, and fill the empty areas with an area less than or equal to the preset area threshold to obtain the boundary of the concave polygon.

[0008] In another possible implementation, the missed job area is determined based on the spatial inclusion mapping between the concave polygon boundary and the initial noise point set, including: In the initial set of noise points, noise points that fall inside the boundary of any concave polygon or whose shortest distance to the nearest concave polygon boundary is less than half of the preset work width are identified as suspected missed work points. The suspected missed work points are filtered for abnormal noise sources to obtain the target missed work data point set; The target set of missing operation data points is subjected to a second preset density clustering process to obtain the missing operation region; wherein, the search neighborhood radius parameter of the second preset density clustering process is smaller than the preset operation width.

[0009] In another possible implementation, the suspected missed work points are filtered for abnormal noise sources to obtain a set of target missed work data points, including: Obtain the timestamp and instantaneous velocity of each suspected missed work point; Among the suspected missed work points, those whose timestamp and the time interval between the adjacent trajectory points are greater than the preset duration threshold, and which do not have other suspected missed work points within the surrounding preset range, are identified as isolated noise points and removed. Points with instantaneous speeds exceeding a preset speed threshold among suspected missed work points are identified as non-work movement noise points and removed. The suspected missed operation points that are retained after being removed are used as the target missed operation data point set.

[0010] In another possible implementation, the preset overlap area constraint is represented as a preset percentage of the preset work width; based on the work trajectory data, the preset work width, and the preset overlap area constraint, the excess overlap area is determined, including: Based on the temporal relationship and heading change characteristics of the trajectory points in the operation trajectory data, the trajectory points in each operation plot are divided into multiple operation rows; Generate a strip buffer with a preset work width for each work row; Calculate the area of ​​the intersection region between the strip buffers of adjacent job rows as the actual overlapping area; The portion of the actual overlapping area that exceeds the reasonable upper limit of overlap determined by the preset overlapping area constraint is defined as the excess overlapping area.

[0011] In another possible implementation, the comprehensive quality evaluation result for the target work area is determined based on the missed work area, the excessive overlap area, and the concave polygon boundary, including: The missed work rate is determined based on the area of ​​the missed work area and the area of ​​the concave polygon boundary. The excess overlap rate is determined based on the area of ​​the excess overlap region and the area of ​​the concave polygon boundary. The area utilization rate is determined based on the area of ​​the concave polygon boundary, the area of ​​the missed work area, and the area of ​​the excessive overlap area. The omission rate, excess overlap rate, and area utilization rate are weighted and integrated to obtain the comprehensive quality evaluation result.

[0012] In another possible implementation, the preset working width and working trajectory data of the agricultural machinery are obtained, including: The original trajectory dataset of agricultural machinery is obtained through positioning devices; Preprocessing operations are performed on the original trajectory dataset to obtain the job trajectory data. The preprocessing operations include: removing abnormal trajectory points and supplementing missing trajectory points. Based on the equipment information of the agricultural machinery, the preset operating width is obtained by querying the preset equipment database.

[0013] Secondly, this application also provides an automatic evaluation device for agricultural machinery operation quality, comprising: an acquisition unit and a processing unit; The acquisition unit is used to acquire the preset working width and working trajectory data of the agricultural machinery; The processing unit is used to perform a first preset density clustering process on the operation trajectory data according to the preset operation width to obtain the target cluster and the initial noise point set. The processing unit is also used to extract the concave hull boundary of the target cluster based on a preset polygon extraction algorithm to obtain the concave hull polygon boundary corresponding to the target work plot; The processing unit is also used to determine the missed operation area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set; The processing unit is also used to determine the excessive overlap area based on the operation trajectory data, the preset operation width, and the preset overlap area constraints. The processing unit is also used to determine the comprehensive quality evaluation result for the target work area based on the missed work area, the excessive overlap area, and the concave polygon boundary.

[0014] Thirdly, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the automatic evaluation method for agricultural machinery operation quality as described above.

[0015] This application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the automatic evaluation method for agricultural machinery operation quality as described above.

[0016] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the automatic evaluation method for agricultural machinery operation quality as described above.

[0017] This application provides an automatic evaluation method, device, and electronic equipment for agricultural machinery operation quality. Through a single density clustering process, it simultaneously generates clusters and noise point sets. The clusters are then used to extract area evaluation benchmarks through concave hull boundary extraction. Noise points are transformed into missed operation areas through spatial mapping and secondary clustering. Combined with overlap analysis based on work rows, it distinguishes between reasonable overlap and excessive overlap. Finally, it outputs a comprehensive quality evaluation result through weighted fusion of multi-dimensional indicators. Therefore, this application requires only two inputs: operation trajectory data and operation width. It eliminates the need for pre-obtaining plot boundaries and installing additional sensors, automatically completing the entire process from plot identification, missed operation location, overlap differentiation to comprehensive scoring. This significantly improves the automation level and accuracy of agricultural machinery operation quality evaluation. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the architecture of the automatic evaluation system for agricultural machinery operation quality provided in this application.

[0020] Figure 2 This is one of the flowcharts of the automatic evaluation method for agricultural machinery operation quality provided in this application.

[0021] Figure 3 This is the second flowchart of the automatic evaluation method for agricultural machinery operation quality provided in this application.

[0022] Figure 4 This is the third flowchart of the automatic evaluation method for agricultural machinery operation quality provided in this application.

[0023] Figure 5 This is the fourth flowchart of the automatic evaluation method for agricultural machinery operation quality provided in this application.

[0024] Figure 6 This is a schematic diagram of noise points mapped to missed work areas provided in this application.

[0025] Figure 7 This is the fifth flowchart of the automatic evaluation method for agricultural machinery operation quality provided in this application.

[0026] Figure 8 This is a schematic diagram of the structure of the automatic evaluation device for agricultural machinery operation quality provided in this application.

[0027] Figure 9 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] The following is combined with Figures 1 to 9 This application describes an automatic evaluation method, device, and electronic equipment for agricultural machinery operation quality.

[0030] Figure 1 This is a schematic diagram of the architecture of an automatic evaluation system for agricultural machinery operation quality provided in this application. The system includes agricultural machinery 101, a positioning device 102 installed on the agricultural machinery, a data processing server 103 that is communicatively connected to the positioning device, and a user terminal 104 that is communicatively connected to the data processing server.

[0031] When performing field operations, the working width of the agricultural machinery 101 can be obtained in at least one of the following ways: automatically read from the agricultural machinery's Controller Area Network (CAN) bus, query from the preset equipment management database according to the agricultural machinery model, or be entered by the operator through the user terminal.

[0032] The positioning device 102 is used to collect trajectory point data of agricultural machinery during operation at a preset sampling frequency. Each trajectory point data includes latitude and longitude coordinates, instantaneous speed, azimuth angle, and timestamp. The positioning device 102 is also used to upload the collected trajectory point data to the data processing server 103 in real time or in batches.

[0033] The data processing server 103 is used to receive trajectory point data and execute the automatic evaluation method for agricultural machinery operation quality provided in this application embodiment. It performs clustering processing, boundary extraction and quality evaluation on the trajectory point data, and generates quality evaluation results including missed operation areas, excessive overlap areas and comprehensive quality scores.

[0034] User terminal 104 is used to display quality evaluation results and a visual map to operators, and to receive input information such as the work area width and evaluation parameter configuration entered by operators.

[0035] In one possible implementation, the positioning device 102 may be a Global Positioning System (GPS) positioning terminal, a Beidou satellite navigation positioning terminal, or other satellite positioning receiving device; the data processing server 103 may be a local server, a cloud server, or an edge computing node; and the user terminal 104 may be a smartphone, a tablet computer, a vehicle display screen, or a desktop computer. This application embodiment does not specifically limit these aspects.

[0036] It should be understood that the implementing entity of the automatic evaluation method for agricultural machinery operation quality provided in this application can be an automatic evaluation device for agricultural machinery operation quality. This automatic evaluation device can be integrated into the data processing server in the aforementioned system, and the data processing server can call its computing resources to execute the automatic evaluation method for agricultural machinery operation quality provided in this application. Alternatively, the automatic evaluation device for agricultural machinery operation quality can also be an independent electronic device, server, or server cluster, as long as the electronic device, server, or server cluster has the functions of trajectory data processing, cluster analysis, boundary extraction, and quality evaluation required to execute the method of this application. This application does not limit the specific physical form of the automatic evaluation device for agricultural machinery operation quality.

[0037] The following embodiments of this application use an example of an automatic evaluation device for agricultural machinery operation quality (hereinafter referred to as the automatic evaluation device) integrated into a data processing server, with the data processing server executing the automatic evaluation method for agricultural machinery operation quality, to illustrate the automatic evaluation method for agricultural machinery operation quality provided in this application. In the embodiments of this application, the data processing server receives the original trajectory dataset from the agricultural machinery positioning device and, based on the preset operation width and the original trajectory dataset, executes the various steps described in the following method embodiments.

[0038] Figure 2 This is one of the flowcharts of an automatic evaluation method for agricultural machinery operation quality provided in this application, such as... Figure 2 As shown, the method includes the following: Step 201: Obtain the preset operating width and operating trajectory data of the agricultural machinery, and perform the first preset density clustering processing on the operating trajectory data according to the preset operating width to obtain the target cluster and the initial noise point set.

[0039] The preset operating width refers to the lateral width covered by the agricultural machinery in a single operation. This parameter can be automatically read through the CAN bus of the agricultural machinery, or queried from the preset equipment management database according to the agricultural machinery model, or entered by the operator. This application does not restrict the method of obtaining it.

[0040] The operation trajectory data is a sequence of trajectory points collected by the positioning equipment on the agricultural machinery according to a preset sampling frequency. Each trajectory point data includes at least latitude and longitude coordinates, instantaneous speed, azimuth angle and timestamp information.

[0041] Optionally, before acquiring the job trajectory data, preprocessing operations can be performed on the original trajectory dataset. Preprocessing operations include removing abnormal trajectory points and supplementing missing trajectory points to eliminate data noise and make the time intervals of trajectory points uniform.

[0042] After acquiring the preset operating width and operating trajectory data of the agricultural machinery, the automatic evaluation device can perform a first preset density clustering process on the operating trajectory data based on the preset operating width to obtain the target cluster and the initial noise point set. Specifically, the first preset density clustering process refers to a clustering analysis process that divides the trajectory points into different sets based on spatial density distribution. The purpose is to group spatially connected trajectory points into the same cluster, while isolated trajectory points with a density lower than the preset condition are marked as noise points.

[0043] In one possible implementation, the automatic evaluation device can determine the search neighborhood radius parameter for the first preset density clustering process based on the preset work area width, and determine the minimum number of neighborhood samples parameter based on the preset desired point density. Subsequently, the automatic evaluation device performs spatial connectivity detection on the work trajectory data based on the above two parameters, aggregating trajectory points that satisfy the density connectivity characteristics into target clusters, and classifying trajectory points that do not satisfy the density connectivity characteristics into an initial noise point set. Each cluster in the target cluster corresponds to a set of trajectory points within an independent work area, and the initial noise point set includes trajectory points that were temporarily excluded from the clusters by the clustering algorithm due to insufficient trajectory point density in their respective areas.

[0044] Therefore, step 201 sets density clustering parameters based on the working width, and then combines the density distribution characteristics of the agricultural machinery itself in the trajectory data to achieve automatic separation of trajectories of multiple plots and preliminary identification of trajectory points in low-density areas.

[0045] Step 202: Extract the concave hull boundary of the target cluster based on the preset polygon extraction algorithm to obtain the concave hull polygon boundary corresponding to the target work plot.

[0046] Understandably, during actual agricultural machinery operations, the outline of the plots being worked on is often not a regular convex polygon, but rather an irregular shape with concave corners, curved edges, and other irregularities. If the convex hull is used as the plot boundary, the concave part of the plot will also be included in the area calculation, leading to a distortion of the area benchmark.

[0047] Therefore, in order to reconstruct the boundary that fits the true outline of the plot from the discrete trajectory points, step 202 uses a preset polygon extraction algorithm to process each target cluster obtained in step 201, and extracts the concave polygon boundary as the benchmark for the working area of ​​each plot.

[0048] Optionally, the automatic evaluation device determines the concave hull control parameters of the preset polygon extraction algorithm based on the preset working width. These concave hull control parameters determine the sensitivity of the boundary to the concave areas of the trajectory point set: if the parameter value is too small, the boundary will be oversensitive, misjudging the tiny gaps caused by random jitter of the trajectory points as plot concavity, resulting in a broken boundary; if the parameter value is too large, the boundary will tend to be convex, causing inaccurate identification of the true concave edge of the plot.

[0049] Meanwhile, considering that the most significant concave features in the agricultural machinery's operating trajectory are the turning area at the edge of the field and the obstacle avoidance area, whose spatial scale is usually within one to two operating widths, the aforementioned concave control parameters are proportionally related to the operating width. This allows the natural concavity of the plot to be preserved while filtering out the interference of random jitter of the trajectory points. Subsequently, the automatic evaluation device inputs the trajectory point set of the target cluster and the concave control parameters into a preset polygon extraction algorithm to generate an initial concave polygon outline.

[0050] Furthermore, for areas with large blank spaces within the plot, such as ponds or buildings, the preset polygon extraction algorithm may generate internal voids. The automatic evaluation device performs void detection on the initial concave polygon outline, retaining voids with an area larger than a preset area threshold as non-operational areas within the plot, and classifying voids with an area smaller than the preset area threshold as algorithm artifacts and filling them, ultimately obtaining the boundary of the concave polygon.

[0051] Step 203: Determine the missed work area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set.

[0052] It should be noted that when agricultural machinery misses certain areas within a plot due to oversight, there are no trajectory points around those areas. Consequently, trajectory points located at the edge of these areas are marked as noise points by the clustering algorithm due to insufficient neighboring points. In other words, the spatial distribution of noise points reflects the location of the missed work area, and there is a spatial inclusion mapping relationship between the two. Based on this, step 203 utilizes the spatial inclusion mapping relationship between noise points and the missed work area to achieve automatic location of the missed work area.

[0053] In one possible implementation, the automatic evaluation device first filters suspected missed work points from the initial set of noise points, and then performs a second density clustering process on the target set of missed work data points to obtain the missed work area.

[0054] For example, for each noise point, it is determined whether it falls inside the boundary of any concave polygon, or whether it is located outside the concave polygon but its shortest distance to the nearest concave boundary is less than half of the preset working width.

[0055] It should be understood that the boundary tolerance distance of half the width is set here to accommodate the boundary error that may be generated at the recessed part of the depressurized area of ​​the depressurized hull, and to avoid the depressurized noise points located at the edge of the depressurized area being missed due to the strict point-polygon inclusion judgment.

[0056] Optionally, after marking the noise points that meet the above conditions as suspected missed work points, the automatic evaluation device further performs abnormal noise source filtering on the suspected missed work points to exclude isolated noise points caused by temporary loss of positioning signals and noise points caused by non-operational movement, thereby obtaining a set of target missed work data points.

[0057] For example, the search neighborhood radius parameter of the second clustering is smaller than the preset job width, ensuring that the clustered noise points are aggregated into a continuous missed job area, while not erroneously merging noise points belonging to different missed job strips.

[0058] This step transforms noise points, which are originally considered interference in cluster analysis, into effective signals carrying spatial information about job quality, thereby enabling the identification of missed job areas without relying on additional sensors.

[0059] Step 204: Determine the excessive overlap area based on the operation trajectory data, preset operation width, and preset overlap area constraints.

[0060] It should be understood that in actual agricultural machinery operations, a small overlap is usually maintained between adjacent work rows to ensure the integrity of the coverage; this overlap is a normal agronomic practice. However, when the overlap is too large, it indicates a deviation in the work path planning or operation execution, constituting repetitive work that should be avoided. To distinguish between the two, step 204 introduces a preset overlap area constraint to decompose the overlapping area of ​​adjacent work rows.

[0061] The preset overlap area constraint is used to define the upper limit of the overlap area that can be considered a reasonable overlap.

[0062] For example, the preset overlap area constraint can be represented as a preset percentage of the preset working width. That is, based on the product of the overlap segment length and the working width, the maximum allowable overlap area on the overlap segment is determined according to this percentage, and the part exceeding the maximum overlap area is identified as excessive overlap.

[0063] In one possible implementation, when the automatic evaluation device performs step 204, it first divides the work area into multiple work rows based on the work trajectory data and the preset work width, and then determines the excessive overlap area based on the multiple work rows and the preset overlap area constraint.

[0064] Optionally, the automatic evaluation device identifies the turning point based on the timestamp sequence and heading change characteristics of the trajectory points in the operation trajectory data, and divides the trajectory points in each plot into boundaries using the turning point as the boundary.

[0065] Optionally, for each work row, the automatic evaluation device generates a strip buffer with a preset work width and calculates the area of ​​the intersection of adjacent work row buffers as the actual overlap area between the two work rows.

[0066] Furthermore, based on the above examples of preset overlap area constraints, the automatic evaluation device determines the reasonable upper limit of overlap area according to the preset overlap area constraints, and then determines the portion of the actual overlap area that exceeds the reasonable upper limit of overlap area as the excess overlap area.

[0067] Therefore, step 204 decomposes the overlapping area into two parts: the reasonable overlap area and the excessive overlap area. Only the excessive overlap area will participate in the calculation of the subsequent comprehensive quality evaluation results.

[0068] Step 205: Determine the comprehensive quality evaluation result for the target work area based on the missed work area, the excessive overlap area, and the concave polygon boundary.

[0069] After obtaining the missed work areas, excessive overlap areas, and concave polygon boundaries through the aforementioned steps, this step conducts a comprehensive quantitative assessment of the work quality of the target work site.

[0070] Specifically, the automatic evaluation device first calculates the missed work rate, excessive overlap rate, and area utilization rate separately. Then, it assigns preset weights to each of these factors and performs a weighted fusion to obtain a comprehensive quality score. Optionally, the specific process is as follows: Step 205a: Determine the missed work rate based on the area of ​​the missed work area and the area of ​​the concave polygon boundary.

[0071] Among them, the missed operation rate directly reflects the proportion of areas that agricultural machinery fails to cover within the target operation plot.

[0072] Optionally, the missed operation rate is the ratio of the total area of ​​the missed operation area to the area of ​​the concave polygon boundary.

[0073] Step 205b: Determine the excess overlap rate based on the area of ​​the excess overlap region and the area of ​​the concave polygon boundary.

[0074] The excess overlap rate excludes reasonable overlaps and only reflects the degree of unnecessary duplicate coverage caused by deviations in work path planning or operation execution.

[0075] Optionally, the excess overlap rate is the ratio of the total area of ​​the excess overlap region to the area of ​​the concave polygon boundary.

[0076] Step 205c: Determine the area utilization rate based on the area of ​​the concave polygon boundary, the area of ​​the missed work area, and the area of ​​the excessive overlap area.

[0077] Among them, the area utilization rate reflects the overall efficiency of agricultural machinery in completing effective coverage within the target operation plot, after deducting the dual impact of missed operations and excessive overlap.

[0078] Optionally, the area of ​​the missed work area and the area of ​​the excessive overlap area are deducted from the area of ​​the concave polygon boundary to obtain the effective work area, and the ratio of the effective work area to the area of ​​the concave polygon boundary is calculated as the area utilization rate.

[0079] Step 205d: Weighted integration of omission rate, excess overlap rate and area utilization rate to obtain comprehensive quality evaluation results.

[0080] It should be understood that missed operations directly lead to missed planting or harvesting of crops, which has the greatest impact on yield and economic benefits, and therefore is given the highest weight; excessive overlap mainly affects operational efficiency and repetitive costs, while area utilization reflects the overall coverage effect. The two are equally important and are given a relatively low weight.

[0081] For example, the weight of the omission rate can be set to 0.4, the weight of the excess overlap rate can be set to 0.3, and the weight of the area utilization rate can be set to 0.3, with the sum of the three weights being 1. The comprehensive quality score is calculated as follows: Comprehensive Quality Score = 0.4 × (1 - Omission Rate) + 0.3 × (1 - Excess Overlap Rate) + 0.3 × Area Utilization Rate. The score ranges from 0 to 1, with a score closer to 1 indicating higher work quality and a score closer to 0 indicating lower work quality.

[0082] It should be noted that the above weight values ​​are only an example and can be adjusted according to the specific job type in actual application: for example, for sowing operations, the losses caused by missed sowing are relatively large, so the weight of the missed operation rate can be further increased; for harvesting operations, the fuel consumption and time costs caused by repeated harvesting are more prominent, so the weight of the excess overlap rate can be appropriately increased.

[0083] In one possible implementation, the final comprehensive quality evaluation result can be output in the form of a visual map. The spatial locations of the missed work areas and the excessive overlap areas are superimposed on the boundary of the concave polygon. The missed work areas and the excessive overlap areas are distinguished by different colors or fill patterns. At the same time, the area statistics and quality score details of each plot can be attached next to the map.

[0084] Through this visualization map, operators can intuitively understand the spatial distribution of work quality, quickly locate concentrated areas of missed work and excessive overlap, and thus carry out targeted supplementary work or work path optimization.

[0085] Based on the above technical solution, this application embodiment generates both clusters and noise point sets through a single density clustering process. The clusters are then extracted using concave hull boundaries to form an area evaluation benchmark. The noise points are transformed into missed operation areas through spatial mapping and secondary clustering. Combined with overlap analysis based on operation rows, reasonable overlap and excessive overlap are distinguished. Finally, a comprehensive quality evaluation result is output under the weighted fusion of multi-dimensional indicators. Therefore, this application requires only two inputs: operation trajectory data and operation width. It eliminates the need for pre-obtaining plot boundaries and installing additional sensors, automatically completing the entire process from plot identification, missed operation location, overlap differentiation to comprehensive scoring. This significantly improves the automation level and accuracy of agricultural machinery operation quality evaluation.

[0086] Figure 3 This is the second flowchart of an automatic evaluation method for agricultural machinery operation quality provided in this application. Figure 3 As shown, in this method, the work trajectory data is subjected to a first preset density clustering process based on a preset work width to obtain the target cluster and an initial noise point set. Specifically, the method includes the following steps: Step 301: Set the search neighborhood radius parameter and the minimum number of neighborhood samples parameter corresponding to the first preset density clustering process.

[0087] Before implementing the first preset density clustering process, the automatic evaluation device needs to determine two key parameters: the search neighborhood radius parameter and the minimum number of samples in the neighborhood parameter. These two parameters together determine the sensitivity of the density clustering process to the spatial distribution of trajectory points and the granularity of the clustering results. Optionally, this can be divided into the following two processes: Step 1: Determine the search neighborhood radius parameter.

[0088] Optionally, the search neighborhood radius parameter corresponding to the first preset density clustering process is determined based on the preset operation width. The search neighborhood radius parameter determines the maximum spatial distance for determining whether two trajectory points belong to the same density connected region in the density clustering process.

[0089] It should be noted that within the same plot of land, agricultural machinery typically travels back and forth along the work rows, with trajectory points distributed in parallel strips. The spatial distance between the closest pair of trajectory points on adjacent work rows is approximately equal to the preset work width. Density clustering is used to group points by propagating along a chain-like path based on density reachability. As long as the jump distance of each step on the path does not exceed the search neighborhood radius parameter, trajectory points that are far apart at both ends of the path can also be grouped into the same cluster. Therefore, the search neighborhood radius parameter only needs to cover the maximum single-step jump distance on the chain-like path, which is the row span distance between adjacent work rows. Considering positioning errors and uneven row spacing, the work width can be appropriately widened.

[0090] For example, the search neighborhood radius parameter can be set to a value between 2 and 3 times the preset work width, with a default value of 2.5 times the preset work width. This ensures that trajectory points on adjacent work rows within the same plot can be connected to meet the density reachability condition, while the spatial distance between different plots is usually much larger than this search neighborhood radius parameter, thus ensuring that trajectory points from different plots are not incorrectly merged.

[0091] Furthermore, when the average sampling interval of the trajectory points deviates from the normal range, the automatic evaluation device can also adaptively adjust the search neighborhood radius parameter according to the average sampling interval of the trajectory points, and constrain the adjusted parameter value within the preset upper limit value, so as to take into account the stability of the clustering effect under different sampling frequencies.

[0092] Step 2: Determine the minimum number of samples in the neighborhood.

[0093] Optionally, the minimum number of neighborhood samples corresponding to the first preset density clustering process is determined based on the preset expected point density. The minimum number of neighborhood samples parameter specifies how many trajectory points must exist within the neighborhood range defined by the search neighborhood radius parameter for the neighborhood center point to be recognized as a core point and participate in density connectivity expansion.

[0094] The minimum number of samples in the neighborhood parameter reflects the expected point density of the normal operating area. That is, under the condition that the agricultural machinery travels at the normal operating speed and the positioning terminal samples at a fixed frequency, the trajectory points in the normal operating area have a relatively stable distribution density in space.

[0095] In one possible implementation, the automatic evaluation device can determine the minimum number of samples in the neighborhood based on the search neighborhood radius parameter and the preset expected point density of the work area. This minimum number of samples is the number of trajectory points that should be included within a circular area with the search neighborhood radius parameter as its radius under normal operating conditions, serving as a reference lower limit. For example, the lower limit of the minimum number of samples in the neighborhood parameter is set to 3 to avoid incorrect clustering due to isolated points accidentally meeting the conditions.

[0096] Step 302: Based on the search neighborhood radius parameter and the minimum number of neighborhood samples parameter, perform spatial connectivity detection on the operation trajectory data.

[0097] Optionally, for each trajectory point, a neighborhood is constructed with that point as the center and the search neighborhood radius parameter as the radius. The number of trajectory points falling within this neighborhood is counted, and then the following judgment is performed: If the number of trajectory points in the neighborhood is greater than or equal to the minimum number of samples in the neighborhood, then the trajectory point is marked as the core point, and density connectivity is extended to other trajectory points in the neighborhood using it as a seed. If the number of trajectory points within the neighborhood is less than the minimum number of samples in the neighborhood, but the trajectory point is located within the neighborhood of a certain core point, then it is marked as a boundary point and assigned to the cluster to which the corresponding core point belongs. If a boundary point is neither a core point nor any core point, it is temporarily marked as an unclassified point and awaits further determination.

[0098] The automatic evaluation device completes the spatial connectivity detection of all trajectory points by traversing point by point and querying neighborhoods until all core points and their density-reachable trajectory points are classified into the corresponding connected regions.

[0099] Step 303: Aggregate trajectory data points that satisfy the density connectivity feature into target clusters.

[0100] After spatial connectivity detection, the automatic evaluation device aggregates trajectory points that meet the density connectivity characteristics into at least one cluster, forming the target cluster.

[0101] Optionally, each cluster corresponds to an independent work plot. This is because trajectory points within the same plot form a continuous high-density area due to the reciprocating operation of agricultural machinery, and are connected to each other through density reachability, thus being classified into the same cluster; while trajectory points between different plots are spatially separated by physical barriers such as field ridges and roads, with distances far exceeding the search neighborhood radius parameter, making it impossible to establish density reachability, and therefore they are classified into different clusters.

[0102] Therefore, the number of target clusters corresponds to the number of independent work plots covered by the trajectory data, and the trajectory points within each cluster are the set of work trajectory points for the corresponding plot.

[0103] Step 304: Classify the trajectory data points that do not meet the density connectivity feature into the initial noise point set.

[0104] Optionally, during the clustering process, trajectory points that are neither marked as core points nor fall within the neighborhood of any core point do not meet the density connectivity characteristics and cannot be assigned to any cluster. The automatic evaluation device uniformly classifies these trajectory points into the initial noise point set.

[0105] Understandably, these noise points are not randomly distributed in space; their generation has clear physical causes. In agricultural machinery operation scenarios, the density of trajectory points in normal operating areas is high, making them easily identified as core points and grouped into clusters. However, near missed operating areas, because the agricultural machinery has not entered these areas, the trajectory points at their edges cannot meet the core point criteria due to insufficient points within their neighborhood, and are thus marked as noise points. Furthermore, in turning areas at the edge of the field, the density of trajectory points may fluctuate due to changes in operating speed and path curvature, generating some noise points.

[0106] Therefore, the initial noise point set includes both noise points caused by missed operations and noise points caused by other factors, providing a data source for subsequent steps to distinguish and utilize these noise points.

[0107] Based on the above technical solution, this embodiment establishes a correlation between the search neighborhood radius parameter of the first preset density clustering process and the working width. This allows adjacent work row trajectory points formed by the reciprocating operation of agricultural machinery within the same plot to be interconnected through density reachability, while trajectory points from different plots are separated due to spatial isolation. This automatically and accurately divides the trajectories of multiple plots without the need for pre-defining plot boundaries. Furthermore, trajectory points that do not meet the density connectivity characteristic are retained as noise points, providing spatial clues for subsequent identification of missed work areas.

[0108] Figure 4 This is the third flowchart of an automatic evaluation method for agricultural machinery operation quality provided in this application. Figure 4 As shown, this method involves extracting the concave hull boundaries of the target clusters based on a preset polygon extraction algorithm to obtain the concave hull polygon boundaries corresponding to the target work area. Specifically, it includes the following steps: Step 401: Determine the concave hull control parameters corresponding to the preset polygon extraction algorithm based on the preset working width.

[0109] Before using a preset polygon extraction algorithm to extract the concave hull boundaries of each target cluster, the automatic evaluation device first needs to determine the concave hull control parameters of the algorithm.

[0110] The preset polygon extraction algorithm is a geometric processing algorithm that extracts a concave boundary that fits the spatial distribution contour of a discrete point set. This algorithm does not require the generated boundary to contain all points within a single convex polygon; instead, it allows the boundary to generate concavities based on local density variations in the point set, thus more accurately approximating the true outer contour of the point set. For an introduction to the concave control parameters, please refer to step 202 above.

[0111] For example, the concave hull control parameter can be a rolling circle radius, which is used for boundary detection outside the point set by a preset polygon extraction algorithm. Specifically, when the rolling circle rolls along the outer edge of the point set, if the distance between two adjacent trajectory points is less than or equal to the diameter of the rolling circle, the rolling circle will pass outside the two points, and the boundary will closely fit the outer edge of the point set; if the distance between two adjacent points is greater than the diameter of the rolling circle, the rolling circle will concave inward into the point set, and the boundary will bend inward accordingly, forming a concave contour corresponding to the sparse region of the point set.

[0112] Optionally, the automatic evaluation device sets the concave control parameter to the product of a preset working width and a preset proportional coefficient. This is because the most prominent concave features in the agricultural machinery's operating trajectory are the field turning area and the obstacle avoidance area. The width of the field turning area is usually within one to two working widths, and the width of the obstacle avoidance area is usually within half to one working width.

[0113] For example, the concave hull control parameter can be set to 1.2 to 1.5 times the preset working width, with a default value of 1.3 times the preset working width.

[0114] Step 402: Input the target cluster into the preset polygon extraction algorithm, and construct the initial concave hull polygon outline by combining the concave hull control parameters.

[0115] After determining the concave hull control parameters, the automatic evaluation device sequentially inputs the trajectory point sets of each target cluster into a preset polygon extraction algorithm. This algorithm uses the concave hull control parameters as the basis for adjusting the boundary fineness, and constructs a concave polygon that fits the outer contour of the point set within the space occupied by the trajectory point set.

[0116] For example, referring to the example in step 401 above, the preset polygon extraction algorithm uses the concave hull control parameter as the radius constraint to roll and construct the boundary outside the point set. Specifically, when the distance between adjacent trajectory points is less than or equal to the concave hull control parameter, the boundary will closely fit the outer edge of the point set; when the distance between adjacent trajectory points is greater than the concave hull control parameter, the boundary will be concave into the point set, forming a concave contour corresponding to the sparse area of ​​the point set.

[0117] Therefore, the initial concave polygon contour generated by the preset polygon extraction algorithm can reflect the spatial density variation of the trajectory point set, forming a natural concave shape in areas where trajectory points are relatively sparse, such as when turning at the edge of the field and avoiding obstacles, while maintaining a smooth boundary in normal working areas where trajectory points are dense.

[0118] Step 403: Process the empty areas inside the initial concave polygon outline to obtain the concave polygon boundary.

[0119] It should be understood that due to potential obstacles such as ponds, houses, and woodlands within the actual land plot, agricultural machinery will traverse these obstacles during operation, resulting in gaps in the trajectory point set within that area. The default polygon extraction algorithm, when processing such trajectory point sets, may generate holes within the initial concave polygon outline. These holes require further evaluation: whether to retain them as genuine non-operational areas or to eliminate them as algorithmic artifacts.

[0120] Optionally, the automatic evaluation device detects the hollow areas inside the initial concave polygon outline, retains the hollow areas with an area greater than a preset area threshold as non-operational areas within the plot, and fills the hollow areas with an area less than or equal to the preset area threshold to obtain the concave polygon boundary.

[0121] Specifically, the automatic evaluation device first detects each closed void region within the initial concave polygon outline and calculates the area of ​​each void region. The area of ​​each void region is then compared with a preset area threshold. If the area of ​​the cavity is greater than the preset area threshold, it is determined that there is a real obstacle or non-working area in the plot corresponding to the cavity, and it is retained and deducted from the total area of ​​the concave boundary in subsequent area calculations. If the area of ​​the hole is less than or equal to the preset area threshold, the hole is determined to be an artifact caused by uneven sampling of trajectory points or deviation of algorithm edge connection. The hole area is then filled and merged into the concave polygon.

[0122] For example, the preset area threshold can be set according to the working width, such as 0.5 times the square of the working width, to adapt to the void determination requirements under different agricultural machinery operation scenarios.

[0123] After the above hole processing, the final concave polygon boundary is obtained.

[0124] Based on the above technical solution, this application embodiment establishes a correlation between the concave hull control parameters and the working width, enabling the preset polygon extraction algorithm to adaptively fit the contour of the trajectory point set formed by agricultural machinery operations with different widths. This forms realistic concave boundaries in areas such as field turns and obstacle avoidance, while filtering out erroneous concaves caused by random jitter of trajectory points. Through the detection and classification of internal voids, it can automatically distinguish between real obstacles and algorithm artifacts within the plot, ensuring the accuracy of the boundary area and providing an area benchmark that is closer to the actual shape of the working plot for subsequent calculations.

[0125] Figure 5 This is the fourth flowchart of an automatic evaluation method for agricultural machinery operation quality provided in this application. Figure 5 As shown, this method determines the missed work area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set, specifically including the following steps: Step 501: Identify suspected missed work points from the initial set of noise points.

[0126] After obtaining the initial set of noise points in step 101, the automatic evaluation device needs to filter out the noise points related to missed operations. The filtering criterion is the spatial relationship between the noise points and the boundary of the concave polygon.

[0127] Optionally, noise points in the initial noise point set that fall inside the boundary of any concave polygon, or whose shortest distance to the nearest concave polygon boundary is less than half the preset work width, are identified as suspected missed work points.

[0128] The two judgment conditions mentioned above are for different situations of missed operations. The first situation is that the missed operation area is located deep inside the plot, and the agricultural machinery has not entered the area. The trajectory points around it are marked as noise points due to insufficient density. These noise points fall directly inside the boundary of the concave polygon, which can be identified by conventional point-polygon inclusion judgment.

[0129] The second scenario involves missed work areas located at the edge of the land parcel. In this case, the hump boundary creates an inward depression at the missed work area, and noise points at the edge of the missed work area may be located just outside this depression. If only strict point and polygon inclusion judgment is used, these noise points will be missed because they are located outside the boundary. To include these noise points near the boundary depression, the automatic evaluation device adds a distance-based supplementary judgment condition in addition to the point and polygon inclusion judgment. Using half of the preset work area width as the distance threshold, noise points within this threshold range to the nearest hump polygon boundary are identified as suspected missed work points. Using half the work area width as this distance threshold can cover the deviation range of the hump boundary at the depression and also avoid misjudging irrelevant noise points far from the land parcel boundary as suspected missed work points.

[0130] Based on the above determination, the automatic evaluation device extracts suspected missed work points from the initial noise point set.

[0131] Step 502: Perform abnormal noise source filtering on suspected missed work points to obtain the target missed work data point set.

[0132] Understandably, in actual operations, not all suspected missed work points are caused by missed work. Factors such as temporary loss of positioning signals and non-operational movement can also lead to a decrease in trajectory point density, thus generating noise points. To improve the accuracy of missed work area identification, the automatic evaluation device needs to perform abnormal noise source filtering on the suspected missed work points obtained in step 501, eliminating noise points generated by factors other than missed work.

[0133] Alternatively, the filtration process can be performed in the following two ways: First, for isolated noise points caused by the loss of positioning signals, the automatic evaluation device acquires and checks the interval between the timestamp of each suspected missed work point and the timestamp of the adjacent trajectory point. If the time interval is greater than the preset duration threshold and there are no other suspected missed work points within the preset range around the suspected missed work point, then the point is determined to be an isolated noise point caused by the temporary loss of positioning signals and is removed.

[0134] Second, for noise points generated by non-operational movement, the automatic evaluation device obtains the instantaneous speed of each suspected missed operation point. If the instantaneous speed is greater than the preset speed threshold, the point is determined to be a noise point generated by non-operational movement such as agricultural machinery transfer and is eliminated.

[0135] After the above filtering process, the remaining suspected missed job points are identified as the target missed job data point set. The noise points in this set have a high confidence level and can be considered to be caused by missed job factors.

[0136] Step 503: Perform a second preset density clustering process on the target missed job data point set to obtain the missed job region. The search neighborhood radius parameter of the second preset density clustering process is smaller than the preset job width.

[0137] It should be noted that although the noise points in the target missed job data point set after filtering are all generated by missed job factors, they are still spatially discrete. Different missed job areas are separated by at least one normal job row and belong to independent spatial units. In order to aggregate noise points belonging to the same missed job area and form a continuous missed job area, the automatic evaluation device performs a second preset density clustering process on the target missed job data point set.

[0138] The clustering object of the second preset density clustering process is the set of target missed job data points after noise source filtering. The purpose is to aggregate the originally discrete noise points into a spatial continuum representing the missed job area. Accordingly, the parameter settings also need to adapt to the spatial distribution characteristics of the missed job area. In this processing, the search neighborhood radius parameter of the second preset density clustering process is smaller than the preset job width because: There must be at least one normal work row between any two adjacent missed work areas, and the spatial distance between their noise point groups must be greater than or equal to the work width. Setting the neighborhood radius parameter of the second clustering to a value smaller than the work width ensures that noise points from different missed work areas are not incorrectly merged into one area.

[0139] Meanwhile, for noise points in the same missed operation area, they are distributed along the edge of the missed operation. Even in areas where the distance between points is too large due to the acceleration of agricultural machinery, as long as the neighborhood radius parameter is less than the operation width while ensuring connectivity, and with an appropriate minimum sample number threshold, noise points belonging to the same missed operation area can still be successfully aggregated. Meanwhile, truly isolated noise points are excluded from the clustering results due to insufficient neighbor points.

[0140] For example, the search neighborhood radius parameter of the second preset density clustering process is specifically set to 0.8 times the preset operation width; the lower limit of the minimum number of neighborhood samples parameter is set to 3, so as to avoid generating incorrect clusters due to individual isolated points accidentally meeting the conditions.

[0141] Therefore, after the second preset density clustering process, each cluster corresponds to a spatially continuous missed operation area. The automatic evaluation device extracts the outline of each missed operation area to obtain its spatial location and boundary.

[0142] For example, Figure 6 This application provides a schematic diagram of a noise point mapping to a missed work area. Figure 6 In the text, parts (a) and (b) correspond to two different cases: Case 1: The concave hull boundary is intact, and the noise point is inside the concave hull.

[0143] At this point, since the noise points are inside the concave hull, they can be directly detected using the point and polygon inclusion judgment, thereby forming a missed operation area through secondary clustering.

[0144] Scenario 2: The noise point is on the outside of the concave bulge, but the shortest distance from the boundary is less than half of the preset working width.

[0145] At this point, the noise point is outside the concave bulge, but if it falls into the close-range detection area that is half the preset working width, it will also be judged as a suspected missed work point, thus forming a missed work area through secondary clustering.

[0146] Based on the above technical solution, this application embodiment establishes a spatial mapping relationship between noise points and missed work areas, transforming clustered noise into missed work location signals, thus overcoming traditional biases. It employs dual spatial filtering to cover both internal and edge missed work scenarios, combines temporal and velocity filtering to eliminate abnormal noise sources, and uses secondary clustering with a neighborhood radius smaller than the work area width to aggregate noise points into continuous missed work areas, avoiding cross-row erroneous merging. The entire process requires no additional sensors, achieving automatic location and quantification of missed work areas solely based on trajectory data.

[0147] Figure 7 This is the fifth flowchart of an automatic evaluation method for agricultural machinery operation quality provided in this application. Figure 7 As shown, this method determines the excessive overlap area based on the operation trajectory data, the preset operation width, and the preset overlap area constraint conditions. Specifically, it includes the following steps: Step 701: Based on the temporal relationship and heading change characteristics of the trajectory points in the operation trajectory data, divide the trajectory points in each operation plot into multiple operation rows.

[0148] It should be understood that agricultural machinery typically uses a reciprocating travel path when operating in the field, turning around at both ends of the work area. Therefore, the complete work trajectory appears in a time series as several segments with roughly parallel directions and alternating opposite directions. To perform overlap analysis between work rows, the automatic evaluation device first needs to divide the trajectory points within each plot into work rows.

[0149] Optionally, the automatic evaluation device calculates the heading angle change between adjacent trajectory points sequentially based on the timestamp order of the trajectory points. When the heading angle change exceeds a preset angle threshold and multiple consecutive trajectory points maintain this trend, it is determined that the agricultural machinery is in a turning state, and this position is marked as a turning point. Using each turning point as a boundary, the trajectory points between adjacent turning points are divided into an independent work row.

[0150] After this division, each work row corresponds to the work path of agricultural machinery crossing the plot in one direction only once, and the travel directions of adjacent work rows are opposite.

[0151] Step 702: Generate a strip buffer with a width equal to the preset work width for each work row.

[0152] Optionally, for each work row containing a sequence of trajectory points, the automatic evaluation device uses the trajectory line as the center line and expands to both sides by half of the preset work width to generate a strip buffer with a total width equal to the preset work width.

[0153] The strip buffer zone spatially represents the work area that agricultural machinery should theoretically cover when traveling along the work row.

[0154] Step 703: Calculate the area of ​​the intersection region between the strip buffers of adjacent work rows as the actual overlapping area.

[0155] After generating the strip buffer for each job row, the automatic evaluation device calculates the overlap between adjacent job rows.

[0156] Optionally, the automatic evaluation device extracts the spatial intersection polygon region of the respective strip buffers of two adjacent work rows in the work row sequence, and calculates the area of ​​the intersection region as the actual overlap area between the two work rows.

[0157] It should be noted that the aforementioned adjacent work rows refer to two work rows that are sequentially adjacent in time and move in opposite directions, rather than work rows that are arbitrarily adjacent in spatial location. This can accurately reflect the actual overlap between adjacent rows during the reciprocating operation of agricultural machinery.

[0158] Step 704: The portion of the actual overlapping area that exceeds the reasonable overlap limit determined by the preset overlapping area constraint is identified as the excess overlapping area.

[0159] It should be noted that a small amount of overlap between adjacent work rows during agricultural machinery operations is a normal agronomic practice to ensure full coverage and should not be considered an operational defect. To distinguish this reasonable overlap from excessive overlap caused by operational deviations, step 704 introduces a preset overlap area constraint for judgment.

[0160] Optionally, the preset overlap area constraint is represented as a preset percentage of the preset working width, which is determined comprehensively based on the positioning accuracy of the agricultural machinery and the agronomic overlap specifications. For example, the preset percentage ranges from 10% to 20%, with a default value of 15%.

[0161] For each overlapping area of ​​adjacent work rows, the automatic evaluation device determines the length of the overlapping segment along the work row direction, and uses the product of the preset percentage, the preset work width, and the length of the overlapping segment as the reasonable upper limit area of ​​overlap. The reasonable upper limit area of ​​overlap is the maximum acceptable overlap area on the overlapping segment.

[0162] The actual overlapping area is compared with the reasonable upper limit area. If the actual overlapping area does not exceed the reasonable upper limit area, the entire overlapping area is considered a reasonable overlap and no excessive overlap occurs. If the actual overlapping area exceeds the reasonable upper limit area, the area corresponding to the excess part is determined as an excessive overlap area.

[0163] Therefore, the total overlapping area of ​​adjacent work rows is decomposed into two parts: reasonable overlap area and excessive overlap area.

[0164] Based on the above technical solutions, this application embodiment transforms discrete trajectory points into continuous coverage areas by dividing the work rows and constructing strip buffers. Based on the preset overlap area constraint, the overlap area of ​​adjacent rows is decomposed into two parts: reasonable overlap and excessive overlap. This allows the overlap rate index to distinguish between normal agronomic overlap and repeated coverage caused by operational errors, providing a more instructive quantitative basis for evaluating work quality.

[0165] The automatic evaluation device for agricultural machinery operation quality provided in this application is described below. This device corresponds to the transportation network topology segmentation method described above. It is understood that the electronic device, in order to achieve the above functions, includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the transportation network topology segmentation method steps described in conjunction with the embodiments disclosed in this application, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0166] For example, Figure 8 This is a schematic diagram of the automatic evaluation device for agricultural machinery operation quality provided in this application, as shown below. Figure 8 As shown, the automatic evaluation device for agricultural machinery operation quality 800 includes the following modules: acquisition unit 801 and processing unit 802.

[0167] The acquisition unit 801 is used to acquire the preset working width and working trajectory data of the agricultural machinery.

[0168] The processing unit 802 is used to perform a first preset density clustering process on the operation trajectory data according to the preset operation width to obtain the target cluster and the initial noise point set.

[0169] The processing unit 802 is also used to extract the concave boundary of the target cluster based on a preset polygon extraction algorithm to obtain the concave polygon boundary corresponding to the target work plot.

[0170] The processing unit 802 is also used to determine the missed operation area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set.

[0171] The processing unit 802 is also used to determine the excessive overlap area based on the operation trajectory data, the preset operation width, and the preset overlap area constraints.

[0172] The processing unit 802 is also used to determine the comprehensive quality evaluation result for the target work area based on the missed work area, the excessive overlap area, and the concave polygon boundary.

[0173] Optionally, the processing unit 802 is further configured to determine the search neighborhood radius parameter corresponding to the first preset density clustering processing based on the preset operation width, and to determine the minimum number of neighborhood samples corresponding to the first preset density clustering processing based on the preset expected point density.

[0174] Optionally, the processing unit 802 is also used to perform spatial connectivity detection on the operation trajectory data based on the search neighborhood radius parameter and the minimum number of neighborhood samples parameter.

[0175] Optionally, the processing unit 802 is also used to aggregate trajectory data points that satisfy density connectivity characteristics into target clusters.

[0176] Optionally, the processing unit 802 is also configured to classify trajectory data points that do not meet the density connectivity feature into an initial noise point set.

[0177] Optionally, the processing unit 802 is also used to determine the concave hull control parameters corresponding to the preset polygon extraction algorithm based on the preset working width.

[0178] Optionally, the processing unit 802 is also used to input the target cluster into a preset polygon extraction algorithm and construct an initial concave hull polygon outline in combination with concave hull control parameters.

[0179] Optionally, the processing unit 802 is also used to detect the hollow areas inside the initial concave polygon outline, retain the hollow areas with an area greater than a preset area threshold as non-operational areas within the plot, and fill the hollow areas with an area less than or equal to the preset area threshold to obtain the concave polygon boundary.

[0180] Optionally, the processing unit 802 is further configured to identify noise points in the initial noise point set that fall inside the boundary of any concave polygon, or whose shortest distance to the nearest concave polygon boundary is less than half of the preset work width, as suspected missed work points.

[0181] Optionally, the processing unit 802 is also used to perform abnormal noise source filtering on suspected missed work points to obtain a set of target missed work data points.

[0182] Optionally, the processing unit 802 is further configured to perform a second preset density clustering process on the target set of missed job data points to obtain the missed job region. The search neighborhood radius parameter of the second preset density clustering process is smaller than the preset job width.

[0183] Optionally, the acquisition module 801 is also used to acquire the timestamp and instantaneous speed of each suspected missed work point.

[0184] Optionally, the processing unit 802 is further configured to identify and remove points among suspected missed work points that have a time interval between their timestamp and adjacent trajectory points that is greater than a preset duration threshold and that do not have other suspected missed work points within a preset range as isolated noise points.

[0185] Optionally, the processing unit 802 is further configured to identify and remove points among suspected missed work points whose instantaneous speed is greater than a preset speed threshold as non-work movement noise points.

[0186] Optionally, the processing unit 802 is also configured to use the suspected missed operation points retained after rejection as the target missed operation data point set.

[0187] Optionally, the processing unit 802 is also used to divide the trajectory points in each work area into multiple work rows according to the temporal relationship and heading change characteristics of the trajectory points in the work trajectory data.

[0188] Optionally, the processing unit 802 is also used to generate a strip buffer with a width of a preset job width for each job row.

[0189] Optionally, the processing unit 802 is also used to calculate the area of ​​the intersection region between the strip buffers of adjacent job rows as the actual overlap area.

[0190] Optionally, the processing unit 802 is further configured to identify the portion of the actual overlapping area that exceeds the reasonable overlap limit area determined by the preset overlapping area constraint as an excessive overlapping area.

[0191] Optionally, the processing unit 802 is also configured to determine the missed operation rate based on the area of ​​the missed operation area and the area of ​​the concave polygon boundary.

[0192] Optionally, the processing unit 802 is also configured to determine the excess overlap rate based on the area of ​​the excess overlap region and the area of ​​the concave polygon boundary.

[0193] Optionally, the processing unit 802 is also configured to determine the area utilization rate based on the area of ​​the concave polygon boundary, the area of ​​the missed operation area, and the area of ​​the excessive overlap area.

[0194] Optionally, the processing unit 802 is also used to perform weighted fusion of the missed work rate, the excessive overlap rate and the area utilization rate to obtain a comprehensive quality evaluation result.

[0195] Optionally, the processing unit 802 is also used to acquire the original trajectory dataset of the agricultural machinery through the positioning device.

[0196] Optionally, the processing unit 802 is also used to perform preprocessing operations on the original trajectory dataset to obtain job trajectory data. The preprocessing operations include: removing abnormal trajectory points and supplementing missing trajectory points.

[0197] Optionally, the processing unit 802 is also used to query and obtain the preset working width from the preset equipment database based on the equipment information of the agricultural machinery.

[0198] Figure 9 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 9 As shown, the electronic device may include a processor 910, a communications interface 920, a memory 930, and a communication bus 940. The processor 910, communications interface 920, and memory 930 communicate with each other via the communication bus 940. The processor 910 can call logical instructions from the memory 930 to execute any of the automatic evaluation methods for agricultural machinery operation quality described above.

[0199] Furthermore, the logical instructions in the aforementioned memory 930 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0200] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute any of the automatic evaluation methods for agricultural machinery operation quality described above.

[0201] In another aspect, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform any of the automatic evaluation methods for agricultural machinery operation quality described above.

[0202] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0203] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0204] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An automatic evaluation method for agricultural machinery operation quality, characterized in that, The method includes: The preset operating width and operating trajectory data of the agricultural machinery are obtained, and the operating trajectory data are subjected to a first preset density clustering process based on the preset operating width to obtain the target cluster and the initial noise point set. Based on a preset polygon extraction algorithm, the concave hull boundary of the target cluster is extracted to obtain the concave hull polygon boundary corresponding to the target work plot. The missed work area is determined based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set; Based on the operation trajectory data, the preset operation width, and the preset overlap area constraint, the excess overlap area is determined; The comprehensive quality evaluation result for the target work site is determined based on the missed work area, the excessive overlap area, and the concave polygon boundary.

2. The method according to claim 1, characterized in that, Based on the preset operation width, the operation trajectory data is subjected to a first preset density clustering process to obtain a target cluster and an initial noise point set, including: The search neighborhood radius parameter corresponding to the first preset density clustering process is determined according to the preset operation width, and the minimum number of neighborhood samples corresponding to the first preset density clustering process is determined according to the preset expected point density. Spatial connectivity detection is performed on the operation trajectory data based on the search neighborhood radius parameter and the minimum number of samples in the neighborhood parameter; Trajectory data points that satisfy the density connectivity characteristic are aggregated into the target cluster; Trajectory data points that do not meet the density connectivity feature are classified into the initial noise point set.

3. The method according to claim 1, characterized in that, Based on a preset polygon extraction algorithm, the concave hull boundary of the target cluster is extracted to obtain the concave polygon boundary corresponding to the target work site, including: Based on the preset working width, determine the concave hull control parameters corresponding to the preset polygon extraction algorithm; The target cluster is input into the preset polygon extraction algorithm, and the initial concave hull polygon outline is constructed in combination with the concave hull control parameters; Detect the empty areas inside the initial concave polygon outline, retain the empty areas with an area greater than a preset area threshold as non-working areas within the plot, and fill the empty areas with an area less than or equal to the preset area threshold to obtain the boundary of the concave polygon.

4. The method according to claim 1, characterized in that, Determining the missed work area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set includes: In the initial set of noise points, noise points that fall inside the boundary of any concave polygon or whose shortest distance to the nearest concave polygon boundary is less than half of the preset work width are identified as suspected missed work points. The suspected missed work points are filtered for abnormal noise sources to obtain the target missed work data point set; The target set of missing job data points is subjected to a second preset density clustering process to obtain the missing job region; wherein, the search neighborhood radius parameter of the second preset density clustering process is smaller than the preset job width.

5. The method according to claim 4, characterized in that, The suspected missed work points are filtered for abnormal noise sources to obtain a set of target missed work data points, including: Obtain the timestamp and instantaneous speed of each of the suspected missed work points; Among the suspected missed work points, those whose time interval between the timestamp and the adjacent trajectory point is greater than a preset duration threshold and which do not have other suspected missed work points within a preset range are identified as isolated noise points and removed. Among the suspected missed work points, those with an instantaneous speed greater than a preset speed threshold are identified as non-work movement noise points and removed. The suspected missed operation points retained after elimination are used as the target missed operation data point set.

6. The method according to claim 1, characterized in that, The preset overlap area constraint is represented as a preset percentage of the preset work width; determining the excess overlap area based on the work trajectory data, the preset work width, and the preset overlap area constraint includes: Based on the temporal relationship and heading change characteristics of the trajectory points in the operation trajectory data, the trajectory points in each operation plot are divided into multiple operation rows; For each of the aforementioned work rows, a strip buffer with a width equal to the preset work width is generated; Calculate the area of ​​the intersection region between the strip buffers of adjacent job rows as the actual overlapping area; The portion of the actual overlapping area that exceeds the reasonable overlap limit determined by the preset overlapping area constraint is defined as the excess overlapping area.

7. The method according to claim 1, characterized in that, The comprehensive quality evaluation result for the target work site is determined based on the missed work area, the excessive overlap area, and the concave polygon boundary, including: The missed work rate is determined based on the area of ​​the missed work area and the area of ​​the concave polygon boundary. The excess overlap rate is determined based on the area of ​​the excess overlap region and the area of ​​the concave polygon boundary. The area utilization rate is determined based on the area of ​​the concave polygon boundary, the area of ​​the missed work area, and the area of ​​the excessive overlap area; The omission rate, the excess overlap rate, and the area utilization rate are weighted and integrated to obtain the comprehensive quality evaluation result.

8. The method according to claim 1, characterized in that, Acquire the preset working width and working trajectory data of agricultural machinery, including: The original trajectory dataset of the agricultural machinery is obtained through positioning devices; Preprocessing operations are performed on the original trajectory dataset to obtain the job trajectory data. The preprocessing operations include: removing abnormal trajectory points and supplementing missing trajectory points. Based on the equipment information of the agricultural machinery, the preset operating width is obtained by querying the preset equipment database.

9. An automatic evaluation device for agricultural machinery operation quality, characterized in that, include: Acquisition unit and processing unit; The acquisition unit is used to acquire the preset working width and working trajectory data of the agricultural machinery; The processing unit is used to perform a first preset density clustering process on the operation trajectory data according to the preset operation width to obtain a target cluster and an initial noise point set. The processing unit is also used to extract the concave hull boundary of the target cluster based on a preset polygon extraction algorithm to obtain the concave hull polygon boundary corresponding to the target work plot; The processing unit is further configured to determine the missed work area based on the spatial inclusion mapping relationship between the concave polygon boundary and the initial noise point set; The processing unit is also used to determine the excessive overlap area based on the operation trajectory data, the preset operation width, and the preset overlap area constraint conditions; The processing unit is further configured to determine a comprehensive quality evaluation result for the target work site based on the missed work area, the excessive overlap area, and the concave polygon boundary.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the automatic evaluation method for agricultural machinery operation quality as described in any one of claims 1 to 8.

Images