A lidar variable prescription fertilization decision system and method

Abstract

The application relates to the technical field of agricultural planting, in particular to a laser radar variable prescription fertilization decision system and method, which comprises the following steps: collecting point clouds, positions and running speeds by a laser radar and a positioning device; forming decision block units according to the point cloud reliability, crop height and crown point density by an edge computing controller; dividing equipment execution fragments along a working path; determining spatial overlap weights and target fertilizer discharge amounts according to a coverage ratio; triggering positions and controlling fertilizer discharge execution mechanisms according to the running speed and response time of the fertilizer discharge execution mechanisms; receiving execution states collected by a state sensor; determining actual fertilizer discharge amounts and updating unexecuted fragments in combination with the running distance. The application maps the decision block units and the equipment execution fragments, so that the demand area corresponds to the working range; through response distance compensation and execution deviation feedback, the prescription switching corresponds to the fertilizer discharge action position, and the remaining instructions reflect the existing fertilizer discharge deviation.

Description

Technical Field

[0001] This invention relates to the field of agricultural planting technology, and in particular to a lidar variable prescription fertilization decision system and method. Background Technology

[0002] The field of agricultural planting technology mainly involves the collection, spatial positioning, data transmission, planting decision-making, and execution of agricultural machinery information related to crops, soil, and farmland operations. It typically involves the coordinated use of sensors, positioning terminals, operation controllers, databases, and fertilization equipment to perform correlation calculations on crop growth, plot location, soil nutrients, and operation records, and to issue fertilization instructions based on prescription data. Among these, the traditional lidar variable prescription fertilization decision-making system and method utilizes lidar to scan the plot and crop canopy, obtaining point cloud data with coordinates, distance, and reflection intensity. This data is then processed by a terminal or server according to plot boundaries, crop height, canopy density, soil sampling records, and set thresholds to generate the fertilization amount and location correspondence for each operation unit. The prescription data is then transmitted to the variable prescription fertilization equipment. Based on the travel position feedback from the positioning device, the equipment controls the opening, rotation speed, or start / stop sequence of the fertilizer dispensing mechanism to complete the fertilization operation, while simultaneously storing the prescription, execution parameters, and operation trajectory.

[0003] Traditional variable-rate fertilization operations correspond crop height and canopy density formed by lidar point clouds to fixed grids or preset zones, and switch the opening and rotation speed of the fertilizer dispensing mechanism according to the positioning coordinates. When the point cloud is sparse or the ground echo is discontinuous, the local structural indicators change with the scanning state, and the prescription boundary lacks correspondence with the operation path and width. The fertilizer dispensing mechanism responds by causing the fertilizer dispensing position to deviate from the prescription position. The execution parameters and operation trajectory are used for recording, and the fertilizer dispensing deviation is difficult to be fed back to the remaining operation area, resulting in inconsistency between the prescription zone and the mechanical execution area. Fertilizer dispensing control is affected by position misalignment and cumulative deviation. Summary of the Invention

[0004] The purpose of this invention is to provide a lidar variable prescription fertilization decision system and method, which constrains the credibility of crop point cloud information collected by lidar, converts the fertilization decision area formed according to crop structure into a mechanical execution area corresponding to the operation path and operation width of the fertilization equipment, and adjusts the control commands of the unexecuted operation area in combination with the response process and actual execution status of the fertilizer discharge execution mechanism, so as to deal with the problems of inconsistency between prescription zoning and mechanical execution range, misalignment between prescription switching position and actual fertilizer discharge position, and accumulation of fertilizer discharge deviation in continuous operation.

[0005] To achieve the above objectives, the present invention provides a lidar variable prescription fertilization decision system, comprising lidar, positioning device, edge computing controller, fertilizer discharge execution mechanism and status sensor; The lidar and the positioning device respectively output lidar point cloud, position and travel speed to the edge computing controller; The edge computing controller determines the reliability of the point cloud based on the number of effective points in the lidar point cloud and the continuity of ground echoes. It extracts ground points and canopy points based on the coordinates, distances, and reflection intensity of the point cloud within each local point cloud region, determines the crop height and canopy point density, forms decision plot units based on the point cloud reliability, crop height, and canopy point density, forms equipment execution segments along the operation path, determines the spatial overlap weight according to the coverage ratio of the equipment execution segments to the decision plot units, and converts the target fertilizer application rate. The edge computing controller determines the fertilizer discharge response distance based on the travel speed and the response time of the fertilizer discharge actuator, converts the target fertilizer discharge amount into a control command for controlling the fertilizer discharge actuator, moves the trigger position of the control command forward according to the fertilizer discharge response distance, and controls the fertilizer discharge actuator. The status sensor collects the execution status of the fertilizer discharge actuator. The edge computing controller determines the actual fertilizer discharge amount based on the execution status and travel distance, calculates the execution deviation, and updates the control instructions of the unexecuted segments according to the spatial overlap weight.

[0006] Furthermore, the edge computing controller divides the lidar point cloud into local point cloud regions according to the plot coordinates, and counts the number of effective points, the length of continuous segments of ground echo along the scanning direction, and the difference in crop height between adjacent scan results for each local point cloud region. When the number of effective points reaches a threshold, the length of continuous segments reaches a threshold, and the difference in crop height does not exceed a consistency threshold, the corresponding local point cloud region is marked as a reliable region. When any of the conditions is not met, the corresponding local point cloud region is marked as a low-reliability region. The crop height and canopy point density of the low-reliability region are corrected based on the crop height and canopy point density of the reliable regions adjacent to the low-reliability region, and the reliability of the point cloud is determined based on the correction result.

[0007] Furthermore, the edge computing controller determines the adjacency relationship between each local point cloud region according to the plot coordinates, and sequentially compares the crop height difference, canopy point density difference, and reliable region markers of adjacent local point cloud regions. When the crop height difference does not exceed the height difference threshold, the canopy point density difference does not exceed the density difference threshold, and both adjacent local point cloud regions are marked as reliable regions, the adjacent local point cloud regions are merged. When any comparison condition is not met, the regional boundary between adjacent local point cloud regions is retained. The adjacency relationship determination and comparison are repeated until there are no adjacent local point cloud regions that meet the merging conditions. The continuous region formed after merging is determined as the decision plot unit.

[0008] Furthermore, the positioning device outputs the position and speed of the fertilization equipment on the working path to the edge computing controller according to the control cycle. The edge computing controller determines the traveling segment based on the position of adjacent control cycles and expands the traveling segment into a strip coverage area according to the preset working width. When the traveling segment crosses the boundary of the decision plot unit, the traveling segment is divided at the intersection of the working path and the boundary of the decision plot unit, and the divided traveling segment and the corresponding strip coverage area are determined as the equipment execution segment. When the traveling segment does not cross the boundary of the decision plot unit, the traveling segment and the corresponding strip coverage area are directly determined as the equipment execution segment.

[0009] Furthermore, the edge computing controller determines the overlapping area between the strip coverage area of ​​each device execution segment and the decision block unit, and uses the area ratio of each overlapping area to the strip coverage area as the spatial overlap weight of the corresponding decision block unit. When a device execution segment covers a decision plot unit, the corresponding spatial overlap weight is set to all weights, and the target fertilization amount of the decision plot unit, which is predetermined based on the point cloud confidence level, crop height, and canopy point density, is used as the target fertilization amount of the corresponding device execution segment. When a device execution segment covers several decision plot units, the predetermined target fertilization amount of each decision plot unit is read, and the target fertilization amounts are weighted and summarized according to the corresponding spatial overlap weights. The weighted summation result is determined as the target fertilization amount of the corresponding device execution segment.

[0010] Furthermore, the edge computing controller reads the travel speed of the current control cycle and the response time of the fertilizer discharge actuator, and determines the fertilizer discharge response distance by multiplying the travel speed by the response time of the fertilizer discharge actuator; it determines the instruction trigger position corresponding to the fertilizer discharge response distance before the starting position of the equipment execution segment along the working path, and converts the target fertilizer discharge amount of the equipment execution segment into the opening instruction, rotation speed instruction, and start / stop instruction of the fertilizer discharge actuator; when the position output by the positioning device reaches the instruction trigger position, it sends the corresponding control instruction to the fertilizer discharge actuator; when the position output by the positioning device does not reach the instruction trigger position, it maintains the control instruction corresponding to the current equipment execution segment.

[0011] Furthermore, when the travel speed difference between adjacent control cycles reaches the speed update threshold, the edge computing controller uses the travel speed of the current control cycle and the response time of the fertilizer discharge actuator to redetermine the fertilizer discharge response distance of the unexecuted equipment segment; adjusts the instruction triggering position of the unexecuted equipment segment along the operation path according to the redetermined fertilizer discharge response distance, and rearranges the sending order of the corresponding control instructions according to the adjusted instruction triggering position; when the travel speed difference between adjacent control cycles does not reach the speed update threshold, the existing fertilizer discharge response distance, instruction triggering position, and control instruction sending order of the unexecuted equipment segment are maintained.

[0012] Furthermore, the status sensor collects the opening degree of the fertilizer discharge port and the rotation speed of the fertilizer discharge shaft of the fertilizer discharge actuator according to the control cycle. The edge computing controller determines the travel distance of the executed segment based on the position change in adjacent control cycles, and determines the actual fertilizer discharge amount of each executed segment based on the pre-stored calibration correspondence between the opening degree of the fertilizer discharge port, the rotation speed of the fertilizer discharge shaft and the amount of fertilizer discharged per unit travel distance. The actual fertilizer discharge amount is compared with the target fertilizer discharge amount of the corresponding device executed segment. If the target fertilizer discharge amount is different from the actual fertilizer discharge amount, the difference is determined as the execution deviation. If the target fertilizer discharge amount is the same as the actual fertilizer discharge amount, the execution deviation is set to zero.

[0013] Furthermore, the edge computing controller determines the decision plot unit associated with the executed segment that generates the execution deviation, filters out unexecuted segments that have spatial overlap weights with the corresponding decision plot unit and have not yet reached the instruction trigger position; allocates the execution deviation according to the proportion of the spatial overlap weight of the unexecuted segment to the sum of the spatial overlap weights obtained from the filtering, and adjusts the target fertilizer discharge amount and control instruction of the unexecuted segment according to the allocation result; when all the unexecuted segments have reached the instruction trigger position, the allocation of the execution deviation is terminated and the actual fertilizer discharge amount and execution deviation of the corresponding executed segment are saved.

[0014] This invention also provides a lidar variable prescription fertilization decision-making method executed by the aforementioned lidar variable prescription fertilization decision-making system, comprising: acquiring a lidar point cloud containing coordinates, distance, and reflection intensity, and the position, operating path, operating width, and travel speed of the fertilization equipment; determining the point cloud reliability based on the number of effective points in the lidar point cloud and the continuity of ground echoes, forming decision plot units based on the point cloud reliability, crop height, and canopy point density, and determining the target fertilization amount; forming equipment execution segments along the operating path, determining the spatial overlap weight based on the coverage ratio of the equipment execution segments to the decision plot units, and converting the target fertilization amount into a target fertilization discharge amount; determining the fertilization discharge response distance based on the travel speed and the response time of the fertilization discharge execution mechanism, moving the trigger position of the control command forward according to the fertilization discharge response distance, and controlling the fertilization discharge execution mechanism; determining the actual fertilization discharge amount based on the execution status and travel distance of the fertilization discharge execution mechanism, calculating the execution deviation between the actual fertilization discharge amount and the target fertilization discharge amount, and allocating the execution deviation to unexecuted segments according to the spatial overlap weight to update the corresponding control command.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the reliability of point clouds is determined by the number of effective points and the continuity of ground echoes. Decision plot units are formed by combining crop height and canopy point density, thus constraining the interference of scanning state changes on plot boundaries and target fertilization amounts. Spatial overlap weights are established according to the coverage ratio of the equipment execution segments to the decision plot units, converting irregular demand areas into target fertilization amounts arranged along the operation path. The control command trigger position is moved forward based on the travel speed and the response time of the fertilization execution mechanism, ensuring that the prescription switching position corresponds to the fertilization action position. The actual fertilization amount is calculated based on the execution status and travel distance, and execution deviations are allocated to unexecuted segments, ensuring that remaining control commands reflect fertilization deviations that have occurred, reducing positional misalignment and cumulative deviations in continuous operations. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is the main flowchart of the lidar variable prescription fertilization decision-making process of the present invention; Figure 2 This is a schematic diagram of the electrical connections and data flow of the system of the present invention; Figure 3 This is a schematic diagram illustrating the point cloud credibility and decision-making plot unit formation of the present invention; Figure 4This is a schematic diagram of the spatial mapping between the decision-making plot unit and the equipment execution segment of the present invention; Figure 5 This is a timing diagram of fertilizer discharge response compensation and execution deviation feedback in this invention. Detailed Implementation

[0018] The following embodiments, based on the protection scope text, invention content and original disclosure materials, illustrate the implementable methods of the lidar variable prescription fertilization decision system and method. In subsequent embodiments, the lidar point cloud refers to a set of points formed by lidar scanning of plots and crop canopies, containing at least coordinates, distance, and reflection intensity; point cloud reliability refers to the state result formed by the edge computing controller based on the number of effective points in the local point cloud, the continuity of ground echoes, and the consistency of adjacent scan results, used to limit whether the local point cloud can directly participate in crop structure analysis; decision plot unit refers to a continuous spatial region formed after merging or retaining boundaries of adjacent local point cloud regions based on point cloud reliability, crop height, and canopy point density; equipment execution segment refers to the control object that forms a strip coverage area along the operation path and combined with the operation width of the fertilization equipment; spatial overlap weight refers to the proportion of the overlap between the strip coverage area of ​​the equipment execution segment and the decision plot unit to the strip coverage area; fertilizer discharge response distance refers to the advance trigger distance of the instruction determined based on the traveling speed of the fertilization equipment and the response time of the fertilizer discharge execution mechanism; execution deviation refers to the difference between the target fertilizer discharge amount and the actual fertilizer discharge amount of the executed segment. The above terms maintain the same meaning in the system embodiments and method-related embodiments. Subsequent embodiments will continue the correspondence between component names and numerical designations established in the protection scope text, accompanying drawings, and original drawings; where the original disclosure did not assign numerical designations to relevant components, no additional numerical designations will be assigned. Numerical designations are used only to identify components and do not indicate quantity, parameters, order, or range.

[0019] Please see Figures 1 to 5This embodiment provides a lidar variable-rate fertilization decision-making system, applied to the operation of fertilization equipment along a predetermined working path for variable-rate fertilization. The system includes a lidar, a positioning device, an edge computing controller, a fertilizer dispensing actuator, and status sensors. The lidar and the edge computing controller establish data connections, as do the positioning device and the edge computing controller. The edge computing controller establishes a control connection with the fertilizer dispensing actuator, and the status sensors establish execution status acquisition and feedback connections with both the fertilizer dispensing actuator and the edge computing controller. During system operation, the lidar and positioning device provide information on the crop spatial structure and the movement status of the fertilization equipment. The edge computing controller converts the crop demand area into equipment execution segments arranged along the working path and sends control commands in advance based on the response process of the fertilizer dispensing actuator. The status sensors return the execution status of the fertilizer dispensing actuator, and the edge computing controller determines the actual fertilizer dispensing amount and corrects any unexecuted control commands.

[0020] The lidar scans the plot and crop canopy. Each point record carries at least spatial coordinates, ranging results, and reflection intensity. Spatial coordinates determine the point record's location within the plot, ranging results characterize the distance between the lidar and the echo location, and reflection intensity preserves echo response information. The lidar point cloud, formed in the scanning sequence, is transmitted to the edge computing controller. The edge computing controller categorizes the lidar point cloud into multiple local point cloud regions based on the plot coordinates, ensuring that point records within the same local point cloud region correspond to the same local spatial range. These local point cloud regions are used for point cloud reliability assessment, crop height extraction, canopy point density statistics, and comparison of adjacent regions; they are not directly equivalent to the final fertilization area.

[0021] The positioning device outputs the position and speed of the fertilizing equipment to the edge computing controller according to the control cycle. Position is used to determine the current progress of the fertilizing equipment on the operating path and the travel segment between adjacent control cycles. Travel speed is used to determine the fertilizer discharge response distance and recalculate the trigger position for subsequent commands when the speed change reaches the update condition. The lidar point cloud and the data output by the positioning device establish a correspondence through plot coordinates and control cycles. Based on this, the edge computing controller determines the spatial relationship between the scanning area, decision plot unit, equipment execution segment, and the current position of the fertilizing equipment. The control cycle is only used to organize data acquisition and control actions and is not limited to a specific duration; the configuration of the control cycle is based on the ability to continuously obtain position, travel speed, and fertilizer discharge execution status.

[0022] The edge computing controller is responsible for determining point cloud reliability, extracting crop structure information, forming decision-making plot units, dividing equipment execution segments, determining spatial overlap weights, converting target fertilizer application rates, determining fertilizer application response distances, generating control commands, and processing execution deviation feedback. The number threshold, continuity threshold, consistency threshold, height difference threshold, density difference threshold, and speed update threshold used by the edge computing controller are provided by pre-configured judgment conditions in the system. Each threshold constrains the number of valid points, the length of continuous ground echo segments, the height difference between adjacent scanned crops, the height difference between crops in adjacent local point cloud regions, the canopy point density difference between adjacent local point cloud regions, and the travel speed difference between adjacent control cycles, respectively. These thresholds are used to form executable comparison boundaries.

[0023] The fertilizer discharging actuator receives control commands from the edge computing controller and adjusts the opening of the fertilizer discharge port, the rotation speed of the fertilizer discharge shaft, or the start / stop status accordingly. Each control command corresponds to the target fertilizer discharge amount for a given segment of the equipment's execution. The edge computing controller, based on the existing control correspondence of the fertilizer discharging actuator, converts the target fertilizer discharge amount into commands for opening the fertilizer discharge port, rotating the fertilizer discharge shaft, or starting / stopping the equipment. Status sensors collect data on the opening of the fertilizer discharge port and the rotation speed of the fertilizer discharge shaft according to the control cycle and feed back the execution status to the edge computing controller. The edge computing controller determines the travel distance by combining the positional changes of the fertilizer application equipment within adjacent control cycles. Then, based on the pre-stored calibration correspondence between the opening of the fertilizer discharge port, the rotation speed of the fertilizer discharge shaft, and the fertilizer discharge amount per unit travel distance, it determines the actual fertilizer discharge amount for the executed segment.

[0024] The system operation process includes the following stages: S1: Acquire and organize lidar point cloud and motion status data. The lidar scans the plot and crop canopy, inputting point records containing spatial coordinates, ranging results, and reflection intensity into the edge computing controller; the positioning device inputs the position and speed of the fertilization equipment during the same operation. The edge computing controller allocates the point records to the corresponding local point cloud areas according to the plot coordinates and saves the order of position and speed according to the control cycle. For the same local point cloud area, the edge computing controller retains the scanning order of the point records to determine the continuity of the ground echo along the scanning direction; for adjacent control cycles, the edge computing controller retains the position changes to form the movement segment of the fertilization equipment.

[0025] S2, Determine the point cloud credibility of local point cloud regions. The edge computing controller counts the number of valid points participating in crop structure analysis in each local point cloud region, and identifies continuous ground echo segments based on echo points distributed along the continuous ground surface in the local point cloud, obtaining the length of the continuous segment along the scanning direction. The edge computing controller also compares the crop height corresponding to the current local point cloud region with the crop height in adjacent scan results to obtain the crop height difference between adjacent scan results. The number of valid points, the length of the continuous segment, and the crop height difference are compared with the quantity threshold, the continuity threshold, and the consistency threshold, respectively. The three comparison results together constitute the credibility conditions of the local point cloud region.

[0026] When the number of valid points reaches a threshold, the length of continuous segments reaches a threshold, and the crop height difference does not exceed a consistency threshold, the edge computing controller marks the corresponding local point cloud region as a reliable region. The crop height and canopy point density of the reliable region are then incorporated into the decision-making plot unit formation process. If any comparison condition is not met, the edge computing controller marks the corresponding local point cloud region as a low-reliability region and stops directly using the original crop height and original canopy point density of the low-reliability region for region boundary determination. The edge computing controller reads the crop height and canopy point density of the reliable regions adjacent to the low-reliability region, corrects the crop height and canopy point density of the low-reliability region based on the continuity relationship of adjacent reliable regions in the plot coordinates, and uses the corrected crop height and canopy point density as structural information for subsequent comparisons of the low-reliability region.

[0027] S3, forming decision-making plot units. The edge computing controller determines the adjacency relationships between local point cloud regions based on plot coordinates, and sequentially compares crop height difference, canopy density difference, and trusted region marking for adjacent local point cloud regions. When the crop height difference between adjacent local point cloud regions does not exceed the height difference threshold, the canopy density difference does not exceed the density difference threshold, and both adjacent local point cloud regions are marked as trusted regions, the edge computing controller merges the adjacent local point cloud regions. The merged contiguous region inherits the coverage area of ​​the participating local point cloud regions in the plot coordinates and serves as the processing object for the next round of adjacency relationship determination.

[0028] When crop height difference exceeds a height difference threshold, canopy point density difference exceeds a density difference threshold, or adjacent local point cloud regions are not simultaneously marked as reliable regions, the edge computing controller retains the regional boundaries between adjacent local point cloud regions. The edge computing controller repeatedly performs adjacency determination, structural difference comparison, and region merging until no adjacent local point cloud regions simultaneously meet the merging conditions. Each merged contiguous region is then defined as a decision plot unit. The edge computing controller determines the target fertilization amount for each decision plot unit based on point cloud reliability, crop height, and canopy point density, ensuring that the target fertilization amount is correlated with the spatial boundaries of the decision plot unit. The target fertilization amount is output from the agronomic decision layer and is not directly used as the immediate control value for the fertilization execution mechanism.

[0029] S4, forming the equipment execution segment. The positioning device outputs the position of the fertilizing equipment on the work path according to the control cycle. The edge computing controller determines the travel segment based on the position of the equipment in adjacent control cycles and extends the travel segment to both sides of the work path as a strip coverage area according to the working width of the fertilizing equipment. The strip coverage area represents the spatial range of fertilizer discharge corresponding to the fertilizing equipment passing through the corresponding travel segment. The edge computing controller spatially compares the strip coverage area with the boundary of the decision plot unit to determine whether the travel segment crosses the boundary of the decision plot unit.

[0030] When a travel segment crosses the boundary of a decision-making plot unit, the edge computing controller divides the travel segment at the intersection of the operation path and the boundary of the decision-making plot unit, and simultaneously divides the corresponding strip coverage area, so that each segmented travel segment has a defined start position, end position, and strip coverage area. The edge computing controller determines the segmented travel segment and its corresponding strip coverage area as the equipment execution segment. When a travel segment does not cross the boundary of a decision-making plot unit, the edge computing controller directly determines the travel segment and its corresponding strip coverage area as the equipment execution segment. Thus, the decision-making plot unit is used to represent the demand area corresponding to the crop structure, and the equipment execution segment is used to represent the control segment executed by the fertilization equipment along the operation path. The two types of spatial objects are formed separately and connected through subsequent spatial overlap weights.

[0031] S5, determine the spatial overlap weight and target fertilizer application rate. The edge computing controller determines the overlap area between the strip coverage area of ​​each device execution segment and each decision plot unit, and determines the spatial overlap weight of the corresponding decision plot unit as the proportion of the area of ​​each overlap area to the corresponding strip coverage area. The spatial overlap weight represents the proportion of the fertilizer application range of the same device execution segment occupied by different decision plot units. The spatial overlap weight processes the spatial coverage relationship between the device execution segment and the decision plot unit, without changing the target fertilizer application rate of the decision plot unit itself.

[0032] When a device execution segment covers only one decision plot unit, the edge computing controller sets the corresponding spatial overlap weight to the full weight and converts the target fertilization amount of the corresponding decision plot unit into the target fertilization amount of the device execution segment. When a device execution segment covers multiple decision plot units, the edge computing controller performs a weighted summation of the target fertilization amounts of each decision plot unit according to the spatial overlap weight, and determines the weighted summation result as the target fertilization amount of the device execution segment. The target fertilization amount is associated with and stored in relation to the start position, end position, and strip coverage area of ​​the device execution segment for use in subsequent control command generation processes.

[0033] S6, determine the fertilizer discharge response distance and trigger the control command. The edge computing controller reads the travel speed and fertilizer discharge actuator response time of the current control cycle, and determines the fertilizer discharge response distance based on the travel distance corresponding to the travel speed within the fertilizer discharge actuator response time. The fertilizer discharge actuator response time represents the response process experienced by the fertilizer discharge actuator from receiving the control command to reaching the corresponding fertilizer discharge port opening, fertilizer discharge shaft speed, or start / stop state. The edge computing controller determines the command trigger position corresponding to the fertilizer discharge response distance along the work path before the start position of the equipment execution segment, and converts the target fertilizer discharge amount of the equipment execution segment into a control command that matches the fertilizer discharge actuator.

[0034] When the position output by the positioning device reaches the command trigger position, the edge computing controller sends a corresponding control command to the fertilizer dispensing actuator. The actuator responds to changes in opening degree, rotation speed, or start / stop status as the fertilizing equipment continues to move, ensuring that the position of the actuator at the target state corresponds to the starting position of the equipment's execution segment. If the position output by the positioning device has not yet reached the command trigger position, the edge computing controller maintains the control command corresponding to the current equipment execution segment and does not prematurely switch to the target fertilizer dispensing amount of the next equipment execution segment. The control commands are sent in the order they are arranged along the work path.

[0035] In each control cycle, the edge computing controller compares the current travel speed with the travel speed of the previous control cycle. When the travel speed difference between adjacent control cycles reaches the speed update threshold, the edge computing controller uses the travel speed of the current control cycle and the response time of the fertilizer discharge actuator to re-determine the fertilizer discharge response distance of the unexecuted equipment execution segment, adjusts the corresponding command triggering position along the work path, and rearranges the sending order of control commands according to the adjusted command triggering position. When the travel speed difference between adjacent control cycles does not reach the speed update threshold, the edge computing controller maintains the existing fertilizer discharge response distance, command triggering position, and control command sending order of the unexecuted equipment execution segment. Dynamic updates only apply to equipment execution segments that have not yet reached the command triggering position, avoiding repeated rewriting of sent control commands during the fertilizer discharge actuator's response process.

[0036] S7, determine the actual fertilizer discharge volume and execution deviation. The status sensor collects the opening degree of the fertilizer discharge port and the rotational speed of the fertilizer discharge shaft of the fertilizer discharge actuator according to the control cycle, and feeds back the execution status, including the collection timing, to the edge computing controller. The edge computing controller determines the travel distance of the executed segment based on the position change output by the positioning device in adjacent control cycles, ensuring that the fertilizer discharge port opening degree, fertilizer discharge shaft rotational speed, and travel distance correspond to the same executed segment. The edge computing controller calls the pre-stored calibration correspondence and determines the actual fertilizer discharge volume of each executed segment according to the correspondence between the fertilizer discharge port opening degree, fertilizer discharge shaft rotational speed, and fertilizer discharge volume per unit travel distance.

[0037] The edge computing controller compares the actual fertilizer discharge volume of the executed segment with the target fertilizer discharge volume of the corresponding device execution segment. When the target fertilizer discharge volume and the actual fertilizer discharge volume are different, the edge computing controller determines the difference between the target fertilizer discharge volume and the actual fertilizer discharge volume as the execution deviation and associates the execution deviation with the executed segment that caused the deviation. When the target fertilizer discharge volume and the actual fertilizer discharge volume are the same, the edge computing controller sets the execution deviation to zero and does not allocate deviations to the device execution segments that have not yet been executed. The actual fertilizer discharge volume, the target fertilizer discharge volume, and the execution deviation maintain a correspondence within the same device execution segment, avoiding the mixing of execution states between different spatial segments.

[0038] S8, Assign execution bias to unexecuted segments. The edge computing controller first determines the decision block unit associated with the executed segment that generated the execution bias, and then filters unexecuted segments from the device execution segments arranged along the operation path that have spatial overlap weights with the corresponding decision block unit and have not yet reached the command trigger position. The filtering criteria simultaneously limit spatial association and control state; device execution segments that have already reached the command trigger position are no longer considered for execution bias assignment.

[0039] The edge computing controller aggregates the spatial overlap weights of the unexecuted segments obtained from the screening process and allocates execution deviations according to the proportion of each unexecuted segment's spatial overlap weight to the total number of screened spatial overlap weights. Based on the allocation results, the edge computing controller adjusts the target fertilizer discharge amount for each unexecuted segment and regenerates the corresponding control instructions based on the adjusted target fertilizer discharge amount. The regenerated control instructions continue to be executed according to the current fertilizer discharge response distance and instruction trigger position, ensuring that the fertilizer discharge amount differences that have already occurred are carried over to the remaining tasks, without adversely affecting the already completed executed segments.

[0040] When there are still unexecuted segments with spatial overlap weights with the corresponding decision plot unit that have not yet reached the command trigger position, the edge computing controller completes the allocation of execution deviations for this period and continues to receive positioning data and execution status for the next control cycle. When all relevant unexecuted segments have reached the command trigger position, the edge computing controller terminates the allocation of corresponding execution deviations and saves the actual fertilizer discharge and execution deviations of the executed segments. The saved results are used to maintain the correspondence between segment states in the same operation process without changing the boundaries of the already formed decision plot units.

[0041] In the system's electrical coordination, the lidar and positioning device respectively generate detection data inputs. The edge computing controller processes the lidar point cloud, position, travel speed, and execution status in the aforementioned order, and generates control command outputs. The fertilizer discharge actuator converts the control commands into physical actions such as the opening degree of the fertilizer discharge port, the rotation speed of the fertilizer discharge shaft, or the start / stop status. The status sensor then converts the execution status corresponding to the physical actions into feedback data. This forms a closed loop of detection data input, spatial decision processing, control signal output, fertilizer discharge action execution, and execution status feedback. The edge computing controller has clear data sources, processing objects, and subsequent call locations for processing point cloud reliability, spatial overlap weight, fertilizer discharge response distance, and execution deviation. There are no isolated control results in the system that are detached from the operation path, decision-making plot unit, or equipment execution segment.

[0042] Please see Figure 1 and Figure 5 This embodiment provides a lidar variable prescription fertilization decision-making method, executed by the aforementioned lidar variable prescription fertilization decision-making system. The method uses lidar point cloud, the location of the fertilization equipment, the operating path, the operating width, and the travel speed as inputs, uses the equipment execution segments arranged along the operating path and the corresponding control commands as intermediate control objects, and uses the execution status and travel distance of the fertilizer dispensing actuator as feedback basis.

[0043] S101 acquires a lidar point cloud containing spatial coordinates, ranging results, and reflection intensity, and obtains the location, operating path, operating width, and travel speed of the fertilization equipment. The lidar point cloud is formed by lidar scanning of the plot and crop canopy, and the location and travel speed are output by the positioning device according to the control cycle.

[0044] S102, the point cloud reliability is determined based on the number of effective points in the lidar point cloud and the continuity of ground echoes. Decision plot units are then formed based on the point cloud reliability, crop height, and canopy point density, and the target fertilizer application rate for each decision plot unit is determined. Point cloud reliability is used to limit whether the structural information of a local point cloud region can be directly compared with adjacent regions. Low-reliability regions are corrected using the crop height and canopy point density of adjacent reliable regions.

[0045] S103: Equipment execution segments are formed along the operation path. The spatial overlap weight is determined based on the coverage ratio of the equipment execution segments to the decision plot units. The target fertilizer application amount of the decision plot units is then converted into the target fertilizer discharge amount of the equipment execution segments according to the spatial overlap weight. The division of equipment execution segments is also based on the location of adjacent control cycles, the operation width, and the boundaries of the decision plot units.

[0046] S104: Determine the fertilizer discharge response distance based on the travel speed and the response time of the fertilizer discharge actuator, move the trigger position of the control command forward along the work path, and control the fertilizer discharge actuator when the position output by the positioning device reaches the trigger position. When the travel speed difference between adjacent control cycles reaches the speed update threshold, re-determine the fertilizer discharge response distance and trigger position for the equipment execution segments that have not yet been executed.

[0047] S105: The actual fertilizer discharge rate is determined based on the opening of the fertilizer discharge port, the rotational speed of the fertilizer discharge shaft, and the travel distance of the fertilizer application equipment. The actual fertilizer discharge rate is compared with the target fertilizer discharge rate to form an execution deviation. This deviation is then allocated to unexecuted segments that have not yet reached the command trigger position according to spatial overlap weights, thereby updating the target fertilizer discharge rate and control command for the unexecuted segments. When all relevant unexecuted segments have reached the command trigger position, the allocation of the corresponding execution deviation is terminated, and the actual fertilizer discharge rate and execution deviation are saved.

[0048] The above-described method-related embodiments and system embodiments use the same point cloud confidence determination, decision-making plot unit formation, equipment execution segment division, spatial overlap weight determination, fertilizer discharge response distance update, and execution deviation feedback rules. The method steps are jointly carried out by data connections, control connections, and feedback connections between the lidar, positioning device, edge computing controller, fertilizer discharge execution mechanism, and status sensors. The control commands output by the method directly act on the fertilizer discharge execution mechanism, and the execution status of the fertilizer discharge execution mechanism returns to the actual fertilizer discharge amount determination process.

[0049] The structural composition, data source, processing order, judgment conditions, spatial relationships, control relationships, and feedback relationships described in the above embodiments are used to illustrate the implementable methods of the technical solutions recorded in this invention. The thresholds in the embodiments are determined by the configuration supported by the technical solutions recorded in the application documents. Component names, data objects, and processing results have consistent meanings throughout the text. Without departing from the technical solutions recorded in this invention and the scope supported by the original disclosure, any equivalent substitutions, equivalent modifications, equivalent combinations, order adjustments, module-to-module replacements, equivalent transformations of field names, equivalent acceptance of the execution subject, or equivalent changes in the carrier form that can be conceived by those skilled in the art should fall within the scope of protection of this patent. Equivalent changes must not extend to unrecorded protected topics, must not alter the substantive correspondence of technical objects through changes in labeling, and must not replace, delete, or weaken the structural features, methodological features, connection relationships, control relationships, data relationships, and temporal relationships necessary for forming point cloud credibility, decision-making plot units, equipment execution segments, spatial overlap weights, fertilizer discharge response distance, and execution deviation feedback.

Claims

1. A lidar variable-prescription fertilization decision-making system, characterized in that, This includes lidar, positioning devices, edge computing controllers, fertilizer discharge actuators, and status sensors; The lidar and the positioning device respectively output lidar point cloud, position and travel speed to the edge computing controller; The edge computing controller determines the reliability of the point cloud based on the number of effective points in the lidar point cloud and the continuity of ground echoes. It extracts ground points and canopy points based on the coordinates, distances, and reflection intensity of the point cloud within each local point cloud region, determines the crop height and canopy point density, forms decision plot units based on the point cloud reliability, crop height, and canopy point density, forms equipment execution segments along the operation path, determines the spatial overlap weight according to the coverage ratio of the equipment execution segments to the decision plot units, and converts the target fertilizer application rate. The edge computing controller determines the fertilizer discharge response distance based on the travel speed and the response time of the fertilizer discharge actuator, converts the target fertilizer discharge amount into a control command for controlling the fertilizer discharge actuator, moves the trigger position of the control command forward according to the fertilizer discharge response distance, and controls the fertilizer discharge actuator. The status sensor collects the execution status of the fertilizer discharge actuator. The edge computing controller determines the actual fertilizer discharge amount based on the execution status and travel distance, calculates the execution deviation, and updates the control instructions of the unexecuted segments according to the spatial overlap weight.

2. The lidar variable prescription fertilization decision system according to claim 1, characterized in that, The edge computing controller divides the lidar point cloud into local point cloud regions according to the plot coordinates, and counts the number of effective points in each local point cloud region, the length of the continuous segment of the ground echo along the scanning direction, and the difference in crop height between adjacent scanning results. When the number of valid points reaches the number threshold, the length of the continuous segment reaches the continuous threshold, and the crop height difference does not exceed the consistency threshold, the corresponding local point cloud region is marked as a reliable region. When any of the judgment conditions are not met, the corresponding local point cloud region is marked as a low-reliability region. The crop height and canopy point density of the low-confidence region are corrected based on the crop height and canopy point density of the confidence region adjacent to the low-confidence region, and the confidence level of the point cloud is determined based on the correction result.

3. The lidar variable prescription fertilization decision system according to claim 2, characterized in that, The edge computing controller determines the adjacency relationship between each local point cloud region according to the plot coordinates, and compares the crop height difference, canopy point density difference and reliable region markers of adjacent local point cloud regions in turn. When the crop height difference does not exceed the height difference threshold, the canopy point density difference does not exceed the density difference threshold, and adjacent local point cloud regions are all marked as reliable regions, adjacent local point cloud regions are merged. When any comparison condition is not met, the region boundary between adjacent local point cloud regions is retained. Repeat the process of determining and comparing adjacent relationships until there are no adjacent local point cloud regions that meet the merging conditions. The continuous region formed after merging is then determined as the decision plot unit.

4. The lidar variable prescription fertilization decision system according to claim 3, characterized in that, The positioning device outputs the position and speed of the fertilization equipment on the working path to the edge computing controller according to the control cycle. The edge computing controller determines the traveling section according to the position of adjacent control cycles and expands the traveling section into a strip coverage area according to the preset working width. When the travel segment crosses the boundary of the decision plot unit, the travel segment is divided at the intersection of the operation path and the boundary of the decision plot unit, and the divided travel segment and the corresponding strip coverage area are determined as the equipment execution segment. When the travel segment does not cross the boundary of the decision plot unit, the travel segment and the corresponding strip coverage area are directly determined as the device execution segment.

5. The lidar variable prescription fertilization decision system according to claim 4, characterized in that, The edge computing controller determines the overlapping area between the strip coverage area of ​​each device execution segment and the decision block unit, and uses the area ratio of each overlapping area to the strip coverage area as the spatial overlap weight of the corresponding decision block unit. When a device execution segment covers a decision plot unit, the corresponding spatial overlap weight is set to all weights, and the target fertilizer application amount of the decision plot unit, which is predetermined based on the point cloud confidence level, crop height, and canopy point density, is used as the target fertilizer application amount of the corresponding device execution segment; when a device execution segment covers several decision plot units, the predetermined target fertilizer application amount of each decision plot unit is read, and the target fertilizer application amounts are weighted and summarized according to the corresponding spatial overlap weights. The weighted summation result is determined as the target fertilizer discharge amount for the corresponding device execution segment.

6. The lidar variable prescription fertilization decision system according to claim 1, characterized in that, The edge computing controller reads the travel speed of the current control cycle and the response time of the fertilizer discharge actuator, and determines the fertilizer discharge response distance by multiplying the travel speed by the response time of the fertilizer discharge actuator. Determine the command trigger position corresponding to the fertilizer discharge response distance before the start position of the equipment execution segment along the operation path, and convert the target fertilizer discharge amount of the equipment execution segment into the opening command, rotation speed command and start / stop command of the fertilizer discharge actuator; When the position output by the positioning device reaches the command trigger position, a corresponding control command is sent to the fertilizer discharge actuator. When the position output by the positioning device does not reach the command trigger position, the control command corresponding to the current device execution segment is maintained.

7. The lidar variable prescription fertilization decision system according to claim 6, characterized in that, When the difference in travel speed between adjacent control cycles reaches the speed update threshold, the edge computing controller uses the travel speed of the current control cycle and the response time of the fertilizer discharge actuator to redetermine the fertilizer discharge response distance of the unexecuted device segment. According to the redefined fertilizer discharge response distance, the command triggering position of the unexecuted equipment execution segment is adjusted along the operation path, and the sending order of the corresponding control commands is rearranged according to the adjusted command triggering position; When the travel speed difference between adjacent control cycles does not reach the speed update threshold, the existing fertilizer discharge response distance, instruction trigger position, and control instruction sending order of the unexecuted device execution segment are maintained.

8. The lidar variable prescription fertilization decision system according to claim 1, characterized in that, The status sensor collects the opening degree of the fertilizer discharge port and the rotation speed of the fertilizer discharge shaft of the fertilizer discharge actuator according to the control cycle. The edge computing controller determines the travel distance of the executed segment based on the position change in adjacent control cycles, and determines the actual fertilizer discharge amount of each executed segment based on the pre-stored calibration correspondence between the opening degree of the fertilizer discharge port, the rotation speed of the fertilizer discharge shaft and the amount of fertilizer discharged per unit travel distance. The actual fertilizer discharge volume is compared with the target fertilizer discharge volume of the corresponding device execution segment. If the target fertilizer discharge volume is different from the actual fertilizer discharge volume, the difference is determined as the execution deviation. If the target fertilizer discharge volume is the same as the actual fertilizer discharge volume, the execution deviation is set to zero.

9. The lidar variable prescription fertilization decision system according to claim 8, characterized in that, The edge computing controller determines the decision plot unit associated with the executed segment that caused the execution deviation, and filters out the unexecuted segments that have spatial overlap weight with the corresponding decision plot unit and have not yet reached the instruction trigger position. The execution deviation is allocated according to the proportion of the spatial overlap weight of the unexecuted segments to the total spatial overlap weight obtained from the screening, and the target fertilizer discharge amount and control command of the unexecuted segments are adjusted according to the allocation result; When all the unexecuted segments have reached the instruction trigger position, the allocation of the execution deviation is terminated and the actual fertilizer discharge and execution deviation of the corresponding executed segments are saved.

10. A lidar variable-prescription fertilization decision-making method, characterized in that, The execution of the lidar variable prescription fertilization decision system according to any one of claims 1-9 includes: Acquire lidar point clouds containing coordinates, distance, and reflection intensity, along with the location, operating path, operating width, and travel speed of the fertilization equipment; The reliability of the point cloud is determined based on the number of effective points in the lidar point cloud and the continuity of ground echoes. Decision plot units are formed based on the reliability of the point cloud, crop height, and canopy point density, and the target fertilization amount is determined. A device execution segment is formed along the operation path. The spatial overlap weight is determined based on the coverage ratio of the device execution segment to the decision plot unit, and the target fertilizer application amount is converted into the target fertilizer discharge amount. The fertilizer discharge response distance is determined based on the travel speed and the response time of the fertilizer discharge actuator. The trigger position of the control command is moved forward according to the fertilizer discharge response distance, and the fertilizer discharge actuator is controlled. The actual amount of fertilizer discharged is determined based on the execution status and travel distance of the fertilizer discharge actuator. The execution deviation between the actual amount of fertilizer discharged and the target amount of fertilizer discharged is calculated, and the execution deviation is allocated to the unexecuted segments according to the spatial overlap weight to update the control instructions.

Images