Energy output dynamic control method and system of high-frequency electric shock weeding equipment

Abstract

The invention provides an energy output dynamic control method and system for high-frequency electric shock weeding equipment, and relates to the technical field of intelligent control, and the method comprises the steps: calibrating a plurality of high-heat response focuses in an energy action region according to temperature rise rate distribution in a thermodynamic response signal; connecting the high thermal response focus to construct a dynamic action range boundary, and adjusting the boundary form along with the real-time thermal diffusion trend; dividing the action range boundary into a plurality of independent energy regulation and control regions according to thermodynamic gradient characteristics; generating a partition energy compensation coefficient based on the thermal damage degree and the energy accumulation state of each regulation and control region; and high-frequency current output is adjusted according to the partition energy compensation coefficient differentiation until the thermodynamic response signal meets the global inactivation condition. According to the invention, energy output parameters can be dynamically adjusted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, in particular to an energy output dynamic control method and system of a high-frequency electric shock weeding device. BACKGROUND

[0002] The present application relates to the technical field of intelligent control, in particular to an energy output dynamic control method and system of a high-frequency electric shock weeding device. SUMMARY

[0003] The technical problem to be solved by the present application is to provide an energy output dynamic control method and system of a high-frequency electric shock weeding device, which can dynamically adjust energy output parameters.

[0004] To solve the above technical problems, the technical solution of the present application is as follows: In a first aspect, an energy output dynamic control method of a high-frequency electric shock weeding device, the method comprising: Step 1: identifying the above-ground morphological characteristics of target weeds through a visual sensor to generate positioning data containing plant size and spatial distribution; Step 2: calculating the voltage amplitude, frequency and action time length parameters of the initial high-frequency current according to a preset mapping relationship based on the plant size in the positioning data; Step 3: executing the initial energy parameters and applying high-frequency current to the weeds, while simultaneously collecting the thermodynamic response signals of the weed tissues in real time; Step 4: calibrating multiple high-heat response foci in the energy action area according to the temperature rise rate distribution in the thermodynamic response signals; Step 5: connecting the high-heat response foci to construct a dynamic action range boundary and adjusting the boundary morphology according to the real-time heat diffusion trend; Step 6: dividing the action range boundary into several independent energy regulation zones according to the thermodynamic gradient characteristics; Step 7: generating a partition energy compensation coefficient based on the thermal damage degree and energy accumulation state of each regulation zone; Step 8: differentially adjusting the high-frequency current output according to the partition energy compensation coefficient difference until the thermodynamic response signals meet the global inactivation condition.

[0005] In a second aspect, an energy output dynamic control system of a high-frequency electric shock weeding device, comprising: An acquisition module for identifying the above-ground morphological characteristics of target weeds through a visual sensor to generate positioning data containing plant size and spatial distribution; A computing module is configured to calculate, based on the plant size in the positioning data, a voltage amplitude, a frequency and an action time length parameter of an initial high-frequency current according to a preset mapping relationship; the initial energy parameter is executed and the high-frequency current is applied to the weeds, and a thermodynamic response signal of the weed tissue is collected in real time; A calibration module is configured to calibrate a plurality of high-heat response focal points in the energy action area according to the temperature rise rate distribution in the thermodynamic response signal; and a dynamic action range boundary is constructed by connecting the high-heat response focal points, and the boundary shape is adjusted according to the real-time heat diffusion trend. A division module is configured to divide the action range boundary into a plurality of independent energy regulation zones according to the thermodynamic gradient characteristics; a partition energy compensation coefficient is generated based on the heat damage degree and the energy accumulation state of each regulation zone; and the high-frequency current output is differentially adjusted according to the partition energy compensation coefficient until the thermodynamic response signal meets the global inactivation condition.

[0006] The above-mentioned scheme of the present application at least includes the following beneficial effects: In terms of precise weed control efficiency, the visual sensor is used to identify the above-ground morphological characteristics of the target weeds to generate accurate positioning data, which covers the plant size and spatial distribution information; based on this, the initial high-frequency current parameters are calculated according to the preset mapping relationship, so that the energy can be accurately applied to the weeds. In the subsequent process, the high-heat response focal points are calibrated according to the temperature rise rate distribution in the thermodynamic response signal, and the dynamic action range boundary is constructed, which enables the energy action area to closely match the actual position of the weeds and the heat diffusion trend, so that the weeds can be precisely attacked, unnecessary interference and damage to the surrounding crops and soil environment are avoided, and the precision of the weed control operation is improved. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 It is a flowchart of the energy output dynamic control method of the high-frequency electric shock weed control equipment provided by the embodiment of the present application.

[0008] Figure 2 It is a schematic diagram of the energy output dynamic control system of the high-frequency electric shock weed control equipment provided by the embodiment of the present application. DETAILED DESCRIPTION

[0009] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0010] As Figure 1As shown, the embodiment of the present application proposes a dynamic energy output control method for high-frequency electric shock weeding equipment, which comprises the following steps: Step 1, identifying the ground shape features of the target weeds through a visual sensor to generate positioning data containing plant size and spatial distribution; Step 2, calculating the voltage amplitude, frequency and action time length parameters of the initial high-frequency current according to the plant size in the positioning data based on a preset mapping relationship; Step 3, executing the initial energy parameters and applying high-frequency current to the weeds, while collecting the thermodynamic response signals of the weeds in real time; Step 4, calibrating multiple high-heat response foci in the energy action area according to the temperature rise rate distribution in the thermodynamic response signals; Step 5, connecting the high-heat response foci to construct a dynamic action range boundary and adjusting the boundary shape according to the real-time heat diffusion trend; Step 6, dividing the action range boundary into several independent energy regulation zones according to the thermodynamic gradient characteristics; Step 7, generating a partition energy compensation coefficient based on the thermal damage degree and energy accumulation state of each regulation zone; Step 8, differentially adjusting the high-frequency current output according to the partition energy compensation coefficient difference until the thermodynamic response signal meets the global inactivation condition.

[0011] In the embodiment of the present application, the weed shape and distribution are locked through visual recognition, the initial energy parameters are matched based on the plant size in step 2, which ensures that the energy output focuses on the weeds from the source, avoids invalid action on the surrounding crops or soil, and reduces environmental interference; the thermodynamic response signals are collected in real time, and steps 4-5 calibrate the high-heat response foci and dynamically adjust the action range boundary, so that the energy action area can be adapted in real time with the heat diffusion of the weeds, ensuring that different parts such as the core area and the crown layer edge of the tall weeds can be effectively covered, solving the problem of incomplete local inactivation under fixed parameters; independent regulation zones are divided according to the thermodynamic gradient, and the compensation coefficient is generated based on the thermal damage degree and energy accumulation state of each region, realizing differential energy output, enhancing energy in areas with insufficient heat response, and reducing output in areas that have met the standard, avoiding excessive energy consumption, and significantly improving the weeding effect per unit of energy consumption; through repeated execution of the "monitoring-analysis-regulation" closed loop, the compensation coefficient is continuously updated and the output is dynamically optimized until the global inactivation condition is met, ensuring that the weeding target can still be stably achieved under complex scenes such as differences in weed growth state and environmental interference, improving the adaptability and reliability of the equipment.

[0012] In a preferred embodiment of the present application, the step 1 comprises: Step 11, synchronously capturing the left-view image and the right-view image of the target area through the calibrated binocular visual sensor; Step 12, respectively, left and right view images are executed on the vegetation spectral response enhancement processing, segmentation of the weed pixel area and generate a binary contour mask; Step 13, based on the binary contour mask extraction feature matching point set, through the binocular disparity calculation to generate depth map; Step 14, the depth map is converted into a three-dimensional space point cloud model, and the ground reference plane is fitted; Step 15, the vertical distance from the highest point of each connected domain in the point cloud model to the ground reference plane is calculated as the plant height, and the minimum circumscribed circle diameter of the point cloud projection in the horizontal plane is calculated as the crown diameter; Step 16, with the point cloud connected domain bottom centroid coordinates as the plant location reference, output the structured positioning data containing three-dimensional coordinates, plant height and crown diameter.

[0013] In the embodiment of the present application, the specific implementation process of step 11 is as follows: First, the calibration of the binocular vision sensor is carried out: the standard chessboard calibration plate is placed in the target shooting area, and it is ensured that the calibration plate completely covers the field of view of the sensor; control the left and right cameras to shoot at least 10 groups of images with different angles and distances, and through the calibration analysis of the pixel coordinates of the chessboard corner points in the image, compare with the actual physical coordinates of the calibration plate, calculate the optical distortion parameters (such as radial distortion, tangential distortion) of the two cameras according to the coordinate differences, and generate a correction matrix to eliminate the distortion of the subsequent shooting images. At the same time, by analyzing the position relationship of the corresponding corner points in the left and right images, the relative position parameters (such as baseline length, horizontal angle, vertical angle) of the two cameras are determined to ensure that the imaging planes of the left and right cameras are accurately aligned in three-dimensional space; after calibration, the hardware synchronization trigger function of the sensor is enabled, and the shutter of the left and right cameras is controlled to open at the same time through the same pulse signal, and the shooting of the target farmland area is completed in the same exposure time (such as 1 / 1000 second), obtaining the left and right view images which are completely synchronized in time and scene.

[0014] The specific implementation process of step 12 is as follows: The left and right perspective images are respectively subjected to vegetation spectral response enhancement processing: first, the color image is converted into a multi-spectral component (such as RGB three channels and a near-infrared channel), the signal of the weed area is strengthened by calculating the ratio of the near-infrared channel to the red light channel (i.e. the principle of the NDVI vegetation index), and the ratio is significantly higher than the soil (low near-infrared reflectivity) and crop residue (no chlorophyll, the ratio is close to the soil) because the weeds contain chlorophyll, the near-infrared reflectivity is high, and the red light reflectivity is low. The histogram equalization processing is performed on the calculated ratio image to stretch the gray value range, so that the gray difference between the weeds and the background is more obvious. Then, the weed pixel segmentation is performed: from the enhanced image, randomly select typical weed and non-weed pixel samples, and count the gray value distribution of the two types of samples to determine the gray threshold value for distinguishing weeds and background (such as the gray value of the weed area is higher than the threshold value, and the non-weed area is lower than the threshold value). According to the threshold value, the image is judged pixel by pixel, and the pixels with a gray value higher than the threshold value are marked as "weed pixels" (assigned a value of 255, white), and the pixels with a gray value lower than the threshold value are marked as "non-weed pixels" (assigned a value of 0, black), and finally a binary contour mask is generated which can completely outline the contour of the weeds.

[0015] The specific implementation process of the above step 13 is as follows: The feature matching point set is extracted from the binary contour mask: in the weed pixel area of the left perspective image, the edge lines of the weed contour are identified by an edge detection algorithm (such as Canny operator), and then the points with sudden curvature (such as corner points) and the points with sudden gray value changes (such as texture junctions) are selected from the edge lines, which are used as feature points of the left image. The same method is used to extract feature points in the weed pixel area of the right perspective image. The local feature descriptors (such as feature vectors based on the gray value distribution of surrounding pixels) of the feature points of the left and right images are calculated, and the similarity between the feature points of the left image and the feature points of the right image is calculated (such as the smaller the vector distance, the higher the similarity), and then a one-to-one corresponding feature point pair is matched (such as the left image feature point A corresponds to the right image feature point A'). According to the principle of binocular vision, the baseline length (calibrated physical distance, such as 10 cm) and focal length (calibrated optical parameters, such as 5 mm) of the left and right cameras are known, and the horizontal pixel coordinate difference (i.e. disparity, such as the left image x coordinate is 100, the right image x coordinate is 80, and the disparity is 20 pixels) of each pair of feature points in the left and right images is calculated. Then, combined with the pixel size (such as 0.01 mm per pixel), the disparity is converted into actual depth: the distance from the feature point to the camera = baseline length x focal length ÷ (disparity x pixel size), and finally a depth map is generated in which each pixel position corresponds to an actual depth.

[0016] The specific implementation process of the above step 14 is as follows: Convert the depth map to a three-dimensional space point cloud model: for each pixel in the depth map, the two-dimensional coordinates (u, v, i.e. horizontal and vertical pixel indices) and the depth value d (the distance from the feature point to the camera) are known. Combined with the principal point coordinates (image center pixel coordinates, such as u0=320, v0=240) and the focal length f (such as 5mm) in the camera intrinsic parameters, the three-dimensional coordinates corresponding to the pixel are calculated through coordinate transformation: X=(u-u0) x d ÷ f, Y=(v-v0) x d ÷ f, Z=d (Z axis is the direction of the camera optical axis). Traverse all pixels in the image, and collect the calculated three-dimensional coordinates (X, Y, Z) to form a three-dimensional point cloud model containing all objects (weeds, soil, crops) in the target area.

[0017] The specific process of fitting the ground reference plane is as follows: First, traverse all points in the three-dimensional point cloud model, extract the Z coordinate (height value along the camera optical axis direction) of each point, and sort all Z coordinates in ascending order. According to the sorting result, select the points with Z value in the lowest 10% interval as the initial candidate points (these points are probably corresponding to the soil surface of the farmland because of their low height). Then check the spatial continuity of these candidate points: calculate the three-dimensional distance between any two candidate points, if the distance is less than a predetermined threshold (such as 10 cm), it is determined to be continuously distributed; remove isolated points that are more than a threshold distance from other candidate points (such points may be small stones, soil protrusions, etc. non-ground features), and finally obtain a preliminary ground candidate point set.

[0018] Remove outliers using RANSAC algorithm: Randomly select 3 non-collinear points (3 points can determine a plane) from the candidate ground point set, take these 3 points as the initial sample, fit a temporary plane through them, calculate the perpendicular distance from all other points in the candidate ground point set to the temporary plane, set a distance threshold (such as 2 cm), if the distance from a point to the plane is less than the threshold, it is determined to be an "inlier" (consistent with the ground plane feature); if the distance is greater than the threshold, it is determined to be an "outlier" (may be a stone in the soil, a protrusion of a weed root, etc. interference points), record the number of inliers corresponding to the current temporary plane, repeat the process of random sampling, fitting a plane, and counting inliers (the number of iterations is usually set to 50-100 times to ensure that enough sample combinations are covered). In all iteration results, select the temporary plane with the most inliers as the optimal preliminary plane, retain all inliers corresponding to the plane, and remove all outliers to obtain a purified ground point set.

[0019] Based on the inlier set (i.e. the purified ground points) filtered by RANSAC, the final ground reference plane is fitted by minimizing the "sum of squared distances of all inliers to the plane": Traverse each point in the inlier set, calculate the perpendicular distance from that point to the plane to be fitted, and sum the squares of all distances to obtain the total error. Adjust the parameters of the plane (a, b, c, d) to gradually reduce the total error until the total error reaches its minimum value (i.e., the overall fit between the plane and all inlier points is the highest). The equation formed by the parameter combination (a, b, c, d) at this point is the equation of the ground reference plane.

[0020] The equation of the ground reference plane is ax + by + cz + d = 0, where the three parameters a, b, and c together constitute the plane's normal vector (a vector perpendicular to the plane), i.e., the normal vector is (a, b, c). The direction of the normal vector determines the plane's tilt angle in three-dimensional space. For example, if a = 0, b = 0, and c = 1, the normal vector is perpendicular to the horizontal plane, and the plane is horizontal (corresponding to flat farmland). If a or b is not 0, the plane is tilted (corresponding to sloping farmland). d represents the vertical distance of the plane from the origin. In short, a, b, and c determine the plane's "orientation" (whether it is tilted and the tilt angle), and d determines the plane's "position" in three-dimensional space (distance from the origin and orientation). Together, they define the specific shape of the ground in three-dimensional space.

[0021] The specific implementation process of step 15 above is as follows: Set a distance threshold between points (e.g., 5 mm), and identify connected components using the region growing method. Starting from any unmarked point, group all adjacent points that are less than the threshold distance into the same region. Repeat this process until all points are marked. Each region is a connected component (representing a single weed or a clump of dense weeds). For each connected component, iterate through the Z-coordinates of all points within it, and record the point corresponding to the maximum Z-value as the highest point of the weed. Calculate the difference between the Z-value of the highest point and the Z-value of the ground reference plane (by substituting the X and Y coordinates of the highest point into the equation of the ground plane to obtain the corresponding Z-value). This difference is the plant height (e.g., if the highest point Z = 20 cm, the corresponding ground Z = 2 cm, and the plant height is 18 cm). When calculating the canopy diameter, ignore the Z-coordinates of all points in the connected component, and only retain the X and Y coordinates to form a set of projection points on a two-dimensional plane. Use the minimum circumcircle algorithm: first find the two points farthest apart in the projection point set as the initial circle diameter, then check if all points are inside the circle. If there are external points, redraw the circle with the point and the farthest point on the circle as the new diameter. Repeat the iteration until all points are inside the circle. At this point, the diameter of the circle is the canopy diameter.

[0022] The specific implementation process of step 16 above is as follows: For each point cloud connected domain, filter out the points with Z value close to the ground reference plane as the bottom point set (e.g. points with a difference of less than 2 cm between Z value and the ground reference Z value). Calculate the centroid coordinates of the bottom point set: add up all the X coordinates of the bottom points and divide by the number of points to get the centroid X coordinate; similarly calculate the centroid Y coordinate and Z coordinate (close to the ground reference Z value), which is the position reference of the weed in three-dimensional space (e.g. X = 1.2 m, Y = 0.8 m, Z = 0.02 m). Finally, integrate the positioning data: create a data entry for each connected domain (weed) containing three-dimensional centroid coordinates (X, Y, Z), plant height (e.g. 18 cm) and crown diameter (e.g. 5 cm) calculated in step 15, output in a structured table format, with table fields including "weed ID", "X coordinate", "Y coordinate", "Z coordinate", "plant height", "crown diameter", to ensure that subsequent steps can directly read these data for energy parameter calculation.

[0023] The application adopts calibrated binocular vision sensors to capture stereo images, providing reliable raw data basis for subsequent three-dimensional reconstruction, ensuring the accuracy reference of spatial positioning; through vegetation spectrum enhancement and binary processing, the weed pixel area is accurately segmented, effectively distinguishing weeds from soil, crops and other backgrounds, reducing the interference of non-target areas, laying a pure image foundation for feature extraction; based on feature matching and disparity calculation, a depth map is generated, and then a three-dimensional point cloud model is constructed and a ground reference is fitted, converting two-dimensional image information into three-dimensional spatial data, realizing the stereoscopic restoration of weed morphology, breaking through the spatial limitations of planar images; through point cloud analysis, the plant height (vertical distance of the highest point) and crown diameter (projected circumscribed circle diameter) are accurately extracted, quantifying the key size features of weeds, providing direct basis for the calculation of initial energy parameters in step 2 (such as height-related voltage, diameter-related action duration); taking the centroid of the connected domain as the position reference, the structured positioning data is output, integrating three-dimensional coordinates, height, diameter and other core information, providing clear spatial coordinate guidance for subsequent energy action range demarcation (such as step 5 boundary adjustment) and precise energy application, ensuring that the energy output strictly matches the actual position and morphology of the weed.

[0024] In a preferred embodiment of the application, step 2 comprises: Step 21, according to the plant height value in the positioning data, referring to the mapping table of preset plant height interval and voltage amplitude reference level, determine the corresponding voltage amplitude reference level; wherein the mapping table is set to correspond to different voltage reference levels when the plant height value falls into different threshold ranges, and the higher the plant height value, the higher the corresponding voltage reference level; Step 22, according to the crown diameter value in the positioning data, the preset crown diameter value and the initial action time positive correlation mapping relationship is calculated to obtain the initial action time; wherein, the positive correlation mapping relationship is set as the larger the crown diameter value is, the longer the initial action time calculated is; Step 23, set a fixed high frequency parameter value, which is configured to be higher than the preset human safety frequency threshold, and lower than the preset electromagnetic interference critical frequency threshold; Step 24, the voltage amplitude reference level determined in step 21 is used as the initial voltage amplitude, the initial action time calculated in step 22 is used as the initial action time, and the fixed high frequency parameter value set in step 23 is combined to generate the voltage amplitude, frequency and action time parameter combination of the initial high frequency current.

[0025] In the embodiment of the application, the specific implementation process of step 21 is as follows: In the laboratory and field environment, a plurality of weed electric shock experiments are carried out: common weed varieties (such as crabgrass, Chinese alpine, gooseweed, etc.) are selected, planted according to the plant height gradient, and typical growth stages such as 0-5 cm, 5-10 cm, 10-15 cm and 15 cm or more are ensured. For each height of weed sample, the voltage amplitude of high frequency current is gradually adjusted (from low to high increment), and the lowest voltage value that can achieve complete inactivation of weed roots is recorded. For example, tests have found that weeds 0-5 cm high can be inactivated at 500 volts, and the highest tolerance is 800 volts (more than that, energy is wasted); weeds 5-10 cm high require 800 volts or more, and can be completely inactivated within 1200 volts, and so on.

[0026] Based on the experimental data, the voltage amplitude reference level is divided: the voltage range that can achieve inactivation without excessive energy consumption is used as the corresponding level of each height interval, and level 1 corresponds to 0-5 cm weeds, the voltage range is 500-800 volts; level 2 corresponds to 5-10 cm weeds, the voltage range is 800-1200 volts; level 3 corresponds to 10-15 cm weeds, the voltage range is 1200-1600 volts; level 4 corresponds to weeds higher than 15 cm, the voltage range is 1600-2000 volts. At the same time, the attribution rule of the interval boundary is marked in the mapping table, such as 5 cm high weeds belonging to level 2 (following the principle of "not low but high", to ensure that higher weeds obtain sufficient initial voltage).

[0027] From the structured positioning data output from step 16, read the "plant height" field value of the current weed to be processed through the data interface. For example, the positioning data of a certain weed records its height as 12 cm, and the system automatically stores this value in a temporary variable; call the mapping table query function to compare 12 cm with the threshold values of each interval in the table one by one: first, compare whether it falls into the 0-5 cm interval (12>5, not matched); then compare the 5-10 cm interval (12>10, not matched); then compare the 10-15 cm interval (10≤12≤15, matched successfully), at this time the system locks the voltage amplitude reference level corresponding to this interval as 3, and extracts the voltage range corresponding to level 3 from the mapping table as 1200-1600 volts.

[0028] Secondly, check the matching result: check whether the height value of the current weed is abnormal (such as whether it exceeds the maximum interval of the mapping table, or whether it is negative), if 12 cm is within a reasonable range (0-∞), then confirm that level 3 and the corresponding voltage range are valid; take 1200-1600 volts as the reference range of the initial voltage amplitude, record it to the energy parameter database at the same time, and mark it as "to be adjusted initial value", provide an initial voltage reference for dynamic optimization in subsequent steps combined with thermodynamic response signals, and generate a log record indicating the current weed height, matched interval, reference level and voltage range, for subsequent tracing and parameter optimization.

[0029] The specific implementation process of the above step 22 is as follows: Select weed samples with different crown diameters (such as 1 cm, 3 cm, 5 cm, 8 cm, 12 cm, etc.), and test the effect of different action durations on weed inactivation under the condition of fixed voltage amplitude and frequency. For example, for weeds with a crown diameter of 1 cm, it is found that 0.7 seconds of action can completely inactivate; weeds with a crown diameter of 3 cm require 1.1 seconds; weeds with a crown diameter of 5 cm require 1.5 seconds, and so on; by analyzing the minimum inactivation duration corresponding to different diameters, determine the three core parameters of "basic action duration", "duration per unit diameter increase" and "maximum action duration". For example, when the crown diameter is 0 cm (theoretically the minimum value), set the basic action duration to 0.5 seconds (to ensure that even the smallest weeds can receive the basic energy); according to the rule "the inactivation duration increases by 0.2 seconds per 1 cm increase in diameter" in the experiment, determine the duration per unit diameter increase as 0.2 seconds / cm; at the same time, combined with energy consumption test, set the maximum action duration not to exceed 5 seconds (to avoid excessive energy application to large crown weeds), record these parameters into the system to form a callable positive correlation mapping relationship database.

[0030] From the structured positioning data output from step 16, the "canopy diameter" value of the current weed is extracted through data field matching. For example, the positioning data of a certain weed records its canopy diameter as 4 cm, and the system stores this value in a dedicated variable; the extracted diameter value is verified for validity: check whether the value is a non-negative number (exclude abnormal data), and whether it is within a reasonable range that the device can handle (such as 0-30 cm, exceeding this range may be a data error). If 4 cm meets the requirements, proceed to the next step of calculation; if the value is abnormal (such as negative or 50 cm), the system automatically triggers an error reporting mechanism, prompting to re-collect data.

[0031] According to the preset positive correlation mapping relationship, the following steps are calculated: The basic duration part is directly taken as the preset basic action duration of 0.5 seconds (corresponding to the benchmark value of canopy diameter 0 cm); the current canopy diameter of 4 cm is multiplied by the time duration of 0.2 seconds / cm for each unit diameter increase, resulting in an incremental duration of 4x0.2=0.8 seconds (i.e. the additional action duration due to the canopy diameter exceeding 0 cm); the basic duration and the incremental duration are added, i.e. 0.5 seconds + 0.8 seconds = 1.3 seconds; 1.3 seconds is compared with the preset maximum action duration of 5 seconds, since 1.3 seconds < 5 seconds, it does not exceed the threshold, so the result is determined as the valid initial action duration. If the extreme case of canopy diameter of 25 cm occurs, the total duration is calculated as 0.5 + 25x0.2 = 5.5 seconds, at this time since it exceeds the maximum threshold of 5 seconds, the system automatically corrects the initial action duration to 5 seconds; the finally determined initial action duration (such as 1.3 seconds) is bound with the ID of the current weed, stored in the energy parameter table, and at the same time the calculation basis of the parameter (including canopy diameter 4 cm, basic duration 0.5 seconds, incremental duration 0.8 seconds) is marked, then the parameter is passed to step 24, combined with the voltage amplitude and frequency parameters to form a complete initial energy output parameter set, providing accurate control basis in time dimension for the subsequent application of high-frequency current.

[0032] The specific implementation process of the above step 23 is as follows: Firstly, two key frequency thresholds are determined: a preset human safety frequency threshold (e.g. 30 kHz, below which the current may cause a risk of electric shock to the human body) and a preset electromagnetic interference critical frequency threshold (e.g. 100 kHz, above which the current may cause strong electromagnetic interference to the surrounding electronic equipment); when setting the high-frequency frequency parameter value, it is necessary to ensure that it is within the safe interval between the two thresholds. For example, through preliminary equipment testing and electromagnetic compatibility experiments, it is determined that 80 kHz is the stable output frequency, which is higher than the human safety frequency threshold of 30 kHz (to avoid safety hazards to the operator) and lower than the electromagnetic interference critical frequency threshold of 100 kHz (to reduce interference to other electronic equipment in the farmland, such as sensors and controllers). 80 kHz is set as the fixed high-frequency frequency parameter value as the frequency reference for the initial energy output of the weeds.

[0033] The specific implementation process of the above step 24 is as follows: The results of steps 21, 22 and 23 are integrated to generate an initial parameter combination: the specific voltage value corresponding to the determined voltage amplitude reference level (e.g. 1400 volts for level 3, which can be an intermediate value within the range of the level voltage or an optimal value selected according to the characteristics of the equipment) obtained from step 21 is taken as the initial voltage amplitude; the initial action duration (e.g. 1.3 seconds) calculated in step 22 is directly taken as the initial action duration parameter; the fixed high-frequency frequency parameter value (e.g. 80 kHz) set in step 23 is taken as the frequency parameter; finally, these three parameters are combined into the initial high-frequency current parameters of "voltage amplitude 1400 volts, frequency 80 kHz, action duration 1.3 seconds", which ensures that the parameter combination can provide basic energy output according to the plant height and crown diameter of the weeds, while meeting the safety and anti-interference requirements, providing an initial reference standard for subsequent dynamic energy adjustment.

[0034] In the embodiments of the present application, step 21 matches the voltage amplitude reference level according to the plant height, so that the voltage output is adapted to the longitudinal growth scale of the weeds, avoiding energy deficiency or waste due to height differences, and ensuring that tall weeds can obtain sufficient initial voltage to reach deep tissues, while low weeds will not be subjected to excessive voltage. Step 22 calculates the initial action duration based on the crown diameter, so that the action time is positively correlated with the lateral coverage of the weeds, and the larger the crown of the weeds, the longer the action time, ensuring that the wide area can fully receive energy, solving the problem of incomplete inactivation in some areas under a uniform time. The fixed high-frequency frequency parameter set in step 23 is both higher than the human safety threshold to ensure operation safety and lower than the electromagnetic interference critical value to reduce the impact on surrounding equipment, achieving a balance between safety and anti-interference; step 24 integrates the parameters to form a complete combination, so that the initial energy output matches the height and crown characteristics of the weeds, and also takes into account the safety and anti-interference requirements, improving energy utilization efficiency and weed targeting from the source.

[0035] In a preferred embodiment of the present application, the step 3 comprises: Step 31, based on the parameter combination of the initial high-frequency current voltage amplitude, frequency and action duration generated in step 24, control the high-frequency current generator to output a high-frequency current pulse sequence with corresponding voltage amplitude, frequency and planned action duration; Step 32, apply the high-frequency current pulse sequence to the target weed plants through the contact electrode array, and the spatial arrangement of the electrode array covers the weed spatial distribution range indicated by the positioning data generated in step 1; Step 33, during the whole process of applying the high-frequency current pulse sequence, use the infrared thermal imager arranged in the gap or adjacent position of the electrode array to continuously capture the infrared thermal radiation image sequence of the target weed plants and the surrounding action area at a sampling rate higher than the frequency of the high-frequency current pulse; Step 34, time series analysis is performed on the infrared thermal radiation image sequence, and the continuous temperature value of each spatial pixel point during the current action is extracted to generate the space-time temperature field data reflecting the real-time temperature change of the weed tissue inside; Step 35, output the space-time temperature field data as a signal representing the thermodynamic response of the weed tissue.

[0036] In the embodiment of the present application, the specific implementation process of the above-mentioned step 31 is as follows: First, extract the key parameters of the initial high-frequency current from the parameter combination generated in step 24: voltage amplitude (such as 1400 volts), frequency (such as 80 kHz) and planned action duration (such as 1.3 seconds). These parameters are transmitted to the control module of the high-frequency current generator through the data interface, and the control module performs legality verification on the parameters to confirm that the voltage amplitude is within the rated output range of the device (such as 500-2000 volts), the frequency meets the pre-set safety interval (such as 30-100 kHz), and the action duration does not exceed the maximum threshold (such as 5 seconds). After the verification is passed, the control module sends instructions to the waveform generation unit inside the generator to set the voltage peak value of the pulse sequence to 1400 volts, the pulse repetition frequency to 80 kHz (i.e. 80000 current pulses per second are output), and the total duration of the pulse sequence to 1.3 seconds through the timer. Before starting the generator, the system performs an open-circuit test to confirm that the voltage and frequency stability of the output pulse meet the error requirements (such as voltage fluctuation ≤±5%, frequency deviation ≤±1 kHz), and the test is passed to output the high-frequency current pulse sequence formally.

[0037] The specific implementation process of the above-mentioned step 32 is as follows: According to the three-dimensional coordinates and spatial distribution range of the weed in the generated positioning data (e.g., a weed is located at X = 1.2 meters, Y = 0.8 meters, and the crown diameter is 4 centimeters), the system controls the mechanical adjustment mechanism to calibrate the spatial arrangement of the contact electrode array. The electrode array is composed of multiple electrodes that can be independently adjusted in position, and the spacing of each electrode is set to be less than 1 / 2 of the crown diameter of the weed (e.g., 2 centimeters), ensuring that the electrodes can uniformly cover the crown area of the weed. By driving the motor to adjust the X and Y axis positions of the electrodes, the center of the electrode array is aligned with the centroid coordinates of the bottom of the weed, and the distribution range of the electrodes completely covers the spatial boundary of the weed indicated by the positioning data (e.g., from X = 1.18 meters to 1.22 meters, Y = 0.78 meters to 0.82 meters). Subsequently, the electrodes are controlled to vertically descend until they come into contact with the top of the weed's crown or the surface of the stem, and the contact pressure is adjusted by the pressure sensor feedback (e.g., maintaining a pressure of 5-10 N to avoid damaging the plant or poor contact), ensuring that the current can be effectively conducted to the weed tissue.

[0038] The specific implementation process of the above step 33 is as follows: An infrared thermal imager is fixed at the gap position of the electrode array (e.g., at the center between adjacent electrodes) or at a position 5 centimeters away from the edge of the electrode, and the lens field of view of the thermal imager is adjusted to completely cover the area within 10 centimeters around the weed plant, ensuring that the complete energy influence range of the current action can be captured. According to the frequency of the high-frequency current pulse in step 31 (80 kHz), the sampling rate of the thermal imager is set to be higher than 100 kHz (i.e., 100000 frames of infrared images per second), ensuring that at least one frame of image records temperature changes within each current pulse period. The temperature measurement range of the thermal imager is set to be from the ambient temperature to 100°C (covering the typical temperature range of weed tissue inactivation), and the temperature measurement accuracy is calibrated to ±0.5°C. At the same time as the high-frequency current pulse sequence starts to output, the thermal imager starts the continuous capture mode synchronously, and each frame of image is stored in the cache area with a timestamp (accurate to microseconds), ensuring that the image sequence is completely synchronized in time with the current action process.

[0039] The specific implementation process of the above step 34 is as follows: The infrared thermal image sequence captured by the infrared thermal imager is imported into the image processing module in chronological order. First, each image is spatially calibrated to determine the correspondence between image pixels and actual spatial positions (e.g., 1 pixel corresponds to an actual distance of 0.1 mm) through a pre-set calibration plate, ensuring that each pixel in the image can be mapped to a specific location in the weed tissue. Time series analysis is performed on the image sequence: the gray value of each pixel in each image is extracted, and the gray value is converted to the corresponding temperature value (e.g., gray value 255 corresponds to 100°C) through the gray-scale-temperature conversion curve of the thermal imager (previously calibrated by black body). The temperature change of each spatial pixel during the entire current action period (1.3 seconds) is tracked, and its temperature value is recorded in chronological order (e.g., a certain pixel is 25°C at 0.1 seconds and 40°C at 0.5 seconds), and finally a space-time temperature field data containing "spatial coordinates (X, Y)-time-temperature" three-dimensional information is generated. The data is stored in matrix form, with each row representing a pixel and each column representing the temperature value at a time point.

[0040] The specific implementation process of step 35 is as follows: The space-time temperature field data generated in step 34 is standardized: sorted by time axis from the start to the end of current action (0 to 1.3 seconds), and the spatial coordinates are remapped according to the actual position of the weed tissue (e.g., a local coordinate system is established with the center of the weed as the origin), ensuring the spatial correlation of the data. Key features such as the highest temperature, average temperature, and temperature change rate at each time point are extracted as core indicators of the thermodynamic response. The standardized space-time temperature field data is packaged as a structured signal, which includes data header (records sampling rate, time length, and spatial range) and data body (temperature matrix), and is transmitted to subsequent steps (e.g., step 4) through internal data bus as the original input signal for analyzing the thermodynamic response of the weed tissue. At the same time, data backup is stored for subsequent process tracing and parameter optimization.

[0041] In the embodiment of the present application, step 31 controls the high-frequency current generator to output corresponding pulse sequences according to the initial parameters, ensuring that the energy is accurately applied according to the preset specifications, and providing stable initial energy input for weed inactivation. Step 32 covers the spatial distribution range of weeds through the electrode array, so that the current action area is accurately matched with the position of the weeds, reducing the diffusion of energy to non-target areas. Step 33 continuously captures thermal radiation images at a high sampling rate, ensuring that the temperature change details of weeds under the action of current can be captured in real time and in detail, avoiding missing key thermodynamic response characteristics. Step 34 extracts continuous temperature values of pixel points through time series analysis to generate spatiotemporal temperature field data, converting abstract thermal response into intuitive data for quantitative analysis. The thermal response signal output by step 35 provides real-time basis for subsequent calibration of high-heat response focal points and dynamic adjustment of energy output, realizing the starting point of the closed loop of "energy application-monitoring-feedback", ensuring that energy regulation can be based on the actual response of weeds, and improving the accuracy and efficiency of weed control.

[0042] In a preferred embodiment of the present application, step 4, according to the temperature rise rate distribution in the thermodynamic response signal, a plurality of high-heat response focal points are calibrated in the energy action area, including: Step 41, based on the spatiotemporal temperature field data output by step 35, calculates the temperature change per unit time of each spatial pixel point in the energy action area within a preset time window, and generates corresponding temperature rise rate field data; Step 42, spatial domain analysis is performed on the temperature rise rate field data, and all pixel points with a temperature rise rate value exceeding a preset first rate threshold are identified to form a candidate high-heat response point set; Step 43, the candidate high-heat response point set is subjected to spatial clustering processing, and candidate points meeting a preset spatial proximity condition are aggregated into several independent high-heat response cluster groups; Step 44, the average temperature rise rate of all points in each high-heat response cluster group is calculated, and cluster groups with an average value exceeding a preset second rate threshold are selected as effective high-heat response cluster groups; Step 45, the spatial center coordinates of each effective high-heat response cluster group are determined, and the center coordinates are calibrated as a high-heat response focal point in the energy action area, and the calibrated high-heat response focal point coordinate set is output.

[0043] In the embodiment of the present application, the specific implementation process of step 41 is as follows: From the spatio-temporal temperature field data output from step 35, extract all spatial pixel coordinates of the energy action area (i.e. the weed and the surrounding area covered by the electrode array) and their corresponding time-temperature sequences (the temperature value of each pixel at different time points); according to the periodic characteristics of high-frequency current pulse (such as pulse frequency 80 kHz, period about 0.0125 milliseconds), set the time window to 5 pulse periods (such as 0.0625 milliseconds), to ensure that a complete temperature change fluctuation can be captured; for each spatial pixel, extract the initial temperature value T1 (at time t1) and the ending temperature value T2 (at time t2) in the continuous preset time window (such as from time t1 to time t2, t2-t1=0.0625 milliseconds) in the time sequence, calculate the temperature change ΔT=T2-T1; divide ΔT by the length of the time window (0.0625 milliseconds) to get the temperature change per unit time of the pixel in the window (i.e. the temperature rise rate, unit: ℃ / millisecond); repeat the above calculation for all pixels in the entire energy action area, and store the temperature rise rate of each point according to its spatial coordinates to generate temperature rise rate field data covering the entire action area (each coordinate point corresponds to a temperature rise rate value).

[0044] The specific implementation process of the above step 42 is as follows: Pre-set first rate threshold: based on the critical temperature rise rate of "irreversible thermal damage of weed tissue begins" in the early experiment (such as the experiment shows that when the temperature rise rate exceeds 0.5 ℃ / millisecond, the weed cells begin to coagulate), the first rate threshold is set to 0.5 ℃ / millisecond; traverse each spatial pixel in the temperature rise rate field data, and compare its temperature rise rate value with the first rate threshold one by one: if the temperature rise rate of a certain pixel is >0.5 ℃ / millisecond, it is determined that the point is a "candidate high heat response point", and its spatial coordinates (such as X=10.2 cm, Y=8.5 cm) are recorded; if ≤0.5 ℃ / millisecond, the point is excluded. The coordinates of all candidate high heat response points that meet the conditions are summarized to form a candidate high heat response point set, ensuring that only the area with rapid response to current action (fast temperature rise) is included in the point set.

[0045] The specific implementation process of the above step 43 is as follows: Set spatial proximity condition: based on the continuity of weed tissue (such as the heat conduction distance between adjacent cells), the preset spatial distance threshold is 0.3 cm (i.e. when the straight-line distance between two points is ≤0.3 cm, it is determined to be spatially adjacent); randomly select an unmarked candidate point from the point set as the initial core point, search for all other candidate points in the point set that are ≤0.3 cm away from the core point, and classify these points and the core point into the same temporary cluster group and mark them as "processed"; for the points in the temporary cluster group that are not core points, repeat the above search process (include unmarked points within a distance of ≤0.3 cm from the point), until the cluster group no longer expands; reselect the core point from the remaining unmarked candidate points, repeat the steps, until all candidate points are marked and classified into the corresponding cluster group, and finally form several independent high-heat response cluster groups, each cluster group representing a continuous high-heat response area (such as a weed growth point or a vascular bundle dense area).

[0046] The specific implementation process of the above step 44 is as follows: Calculate the average rate of temperature rise of each cluster group: for each high-heat response cluster group, extract the rate of temperature rise values of all candidate points in it, add these values and divide by the number of points included in the cluster group (i.e. the total number of points), to obtain the average rate of temperature rise of the cluster group. For example, a cluster group contains 5 points, the rates are 0.6, 0.7, 0.8, 0.7, and 0.6°C / msec, and the average value is (0.6+0.7+0.8+0.7+0.6) ÷ 5 = 0.68°C / msec; in order to ensure that the cluster group as a whole has significant high-heat response (excluding sporadic high-speed point aggregation), set the second rate threshold to 0.6°C / msec (higher than the first threshold, reflecting the high-heat characteristics of the cluster group level); compare the average rate of temperature rise of each cluster group with the second threshold, if the average value > 0.6°C / msec, it is determined to be an "effective high-heat response cluster group"; if ≤0.6°C / msec, it is determined to be an invalid cluster group and is discarded, for example, the above-mentioned cluster group with an average of 0.68°C / msec is retained, while the cluster group with an average of 0.55°C / msec is discarded.

[0047] The specific implementation process of the above step 45 is as follows: For each effective high-heat response cluster, the spatial coordinates of all points in the cluster are extracted, i.e. (X1, Y1), (X2, Y2), …, (Xn, Yn), and the average value of the X coordinates (X center = (X1+X2+…+Xn) / n) and the average value of the Y coordinates (Y center = (Y1+Y2+…+Yn) / n) are calculated to obtain the spatial center coordinates (X center, Y center) of the cluster; the spatial center coordinates of each effective cluster are defined as a "high-heat response focus" representing the core position of the cluster (i.e. the region with the most concentrated heat response in the weed tissue); the center coordinates of all effective clusters are collected to form a high-heat response focus coordinate set (such as [(10.3cm, 8.6cm), (12.1cm, 9.2cm), …]), and the set is output to step 5 as the core reference point for constructing the dynamic action range boundary.

[0048] The step 41 of the present application calculates the global temperature rise rate field, quantifies the heat dynamic change of the weed tissue as spatial distribution data, provides a comprehensive basis for identifying the high-heat response region, and avoids missing the key reaction region; the step 42 accurately locks the region sensitive to the current and with rapid temperature rise by screening the candidate points through the first rate threshold, excludes the low-response background interference, and focuses on the potential inactivation key area; the step 43 highlights the continuously distributed high-heat response region by aggregating the adjacent candidate points into clusters, eliminates the influence of single-point noise, and is more in line with the actual heat reaction characteristics of the weed tissue; the step 44 ensures that the retained cluster has overall significant high-heat response by screening the effective cluster through the second rate threshold, further improves the reliability of the focus calibration, and avoids the interference of atypical regions; the high-heat response focus determined in the step 45 accurately locates the core region with the most severe heat reaction in the weed tissue, provides a clear target for subsequent dynamic adjustment of the energy action range, makes the energy regulation more focused on the key inactivation point, and improves the herbicidal efficiency and accuracy.

[0049] In a preferred embodiment of the present application, the step 5 comprises: Step 51: based on the set of all calibrated high-heat response focus coordinates output in step 45, a spatial topology connection operation is performed to connect adjacent foci that meet the preset distance threshold condition two by two to form an initial polygon boundary network; Step 52: boundary smoothing processing is performed on the initial polygon boundary network to eliminate local sharp corners and generate a continuous closed initial action range boundary contour; Step 53: based on the real-time updated spatiotemporal temperature field data output in step 35, heat diffusion rate field data of the current time in the energy action area are calculated; the heat diffusion rate field data reflect the spatial gradient characteristics of heat propagation from the high-heat area to the low-temperature area; Step 54, according to the thermal diffusion rate field data, predicting the anisotropic thermal diffusion trend within a preset distance from the outer edge of the initial action range boundary contour; wherein the area with high thermal diffusion rate is predicted as the boundary expansion direction, and the area with low thermal diffusion rate is predicted as the boundary maintenance or contraction direction; Step 55, according to the anisotropic thermal diffusion trend predicted in step 54, performing real-time deformation adjustment on the initial action range boundary contour: adjusting the boundary to extend outward by a preset deformation amount in the predicted expansion direction, and keeping or contracting the boundary inward in the predicted maintenance or contraction direction; generating a dynamically updated action range boundary shape, and outputting the action range boundary shape data after dynamic adjustment at the current time.

[0050] In the embodiment of the present application, the specific implementation process of the above step 51 is as follows: From the high-heat-response focus coordinate set output in step 45, extract the three-dimensional space coordinates of all focus points (such as (X1, Y1, Z1), (X2, Y2, Z2) … (Xn, Yn, Zn)), and project them onto the horizontal plane (ignore the Z coordinate, and keep the X and Y coordinates) for planar topological analysis; based on the average density of the weed canopy and the focus point distribution characteristics, set the distance threshold of adjacent focus points to 5 centimeters (i.e. when the straight-line distance between two focus points is ≤5 centimeters, it is determined as "adjacent focus points"); calculate the Euclidean distance between any two focus points in the set, and compare the distance of each pair of focus points with the threshold; if the distance ≤5 centimeters, it is determined that the adjacent condition is met, and the two focus points are connected by a straight line; if the distance >5 centimeters, they are not connected. After traversing all pairs of focus points, an initial polygon boundary network (such as irregular polygon combinations of triangles, quadrilaterals, etc.) composed of multiple line segments is formed, ensuring that the network covers all focus points and the line segments do not intersect.

[0051] The specific implementation process of the above step 52 is as follows: For each corner (i.e. the intersection of three or more line segments) in the initial polygon boundary network, identify the local sharp corner (such as a vertex with an angle <90°); for each sharp corner, take the midpoint of the two adjacent line segments that form the corner, and calculate the midpoint coordinates (such as the line segment endpoints A(Xa, Ya) and B(Xb, Yb), and the midpoint M[(Xa+Xb) / 2, (Ya+Yb) / 2]); replace the original corner vertex with the midpoint of the connecting line of the adjacent segment midpoints, or fit the adjacent line segments by a Bezier curve to make the line segment at the corner transition smooth (such as adjusting the angle to 120°-150°), and repeat the above smoothing operation for the entire boundary network until all corners meet the preset smoothing requirement (such as an angle ≥100°), and finally form a continuous and closed initial action range boundary contour (i.e. a polygon contour without breakpoints and sharp protrusions).

[0052] The specific implementation process of the above step 53 is as follows: From the real-time updated spatio-temporal temperature field data output from step 35, the temperature values of all spatial pixel points in the energy action area at the current time (such as time t) are extracted to form the current temperature field distribution (each pixel point corresponds to a temperature value T(x, y)); for any two adjacent pixel points (such as pixel P(x1, y1) and pixel Q(x2, y2)) in the temperature field, the temperature difference AT = |T(x1, y1) - T(x2, y2)| is calculated, and the spatial distance d between the two points (in centimeters) is measured. The heat diffusion rate v = AT / d (in ℃ / cm), and the rate direction is from the high-temperature pixel to the low-temperature pixel (i.e. the heat propagation direction); all adjacent pixel pairs in the entire energy action area are traversed, and the heat diffusion rate (including size and direction) of each pixel point is stored according to its spatial coordinates to generate heat diffusion rate field data. This data reflects the spatial gradient characteristics of heat propagation from high-temperature areas (such as near the high-heat response focus) to low-temperature areas (such as the edge of the weed or the soil) in vector form.

[0053] The specific implementation process of the above step 54 is as follows: The outer edge preset distance is set to 2 centimeters, that is, a ring-shaped area extending 2 centimeters outward from the boundary contour is set as the target area for heat diffusion trend prediction; the heat diffusion rates of all pixel points in the area are counted, and a rate threshold (such as 0.5 ℃ / cm) is set. The area with a rate > 0.5 ℃ / cm is determined as a "high heat diffusion rate area", and the area with a rate ≤ 0.5 ℃ / cm is determined as a "low heat diffusion rate area"; the high heat diffusion rate area indicates that the weed tissue in this direction is still rapidly absorbing heat due to fast heat propagation, so the energy action range needs to be expanded to cover the diffusion area, and is predicted as a "boundary expansion direction"; the low heat diffusion rate area indicates that the heat propagation is slow, indicating that it is close to energy saturation or there is no more weed tissue, so it is predicted as a "boundary maintenance direction" (when the rate is close to the threshold) or a "boundary contraction direction" (when the rate is much lower than the threshold).

[0054] The specific implementation process of the above step 55 is as follows: According to the prediction result of step 54, the boundary adjustment parameter is set: in the expansion direction, the extension amount is determined according to the size of the thermal diffusion rate (for example, the extension amount increases by 0.3 cm for every 0.1 ℃ / cm increase in the rate, and the maximum extension amount does not exceed 1 cm); in the maintenance direction, the boundary position is kept unchanged; in the contraction direction, the contraction amount is determined according to the degree of rate reduction (for example, the contraction amount increases by 0.2 cm for every 0.1 ℃ / cm reduction in the rate, and the maximum contraction amount does not exceed 0.5 cm); each vertex of the boundary contour is traversed, and the prediction trend in the direction of the vertex is judged; if it is the expansion direction, the vertex is moved along the thermal diffusion direction by the corresponding extension amount; if it is the maintenance direction, the vertex position is unchanged; if it is the contraction direction, the vertex is moved to the inside of the contour by the corresponding contraction amount; after adjustment, all the vertices are connected again to form a new closed contour, a dynamically updated action range boundary form is generated, and the boundary form data (including contour vertex coordinates, boundary length, etc.) at the current time is output, so as to ensure that the boundary always fits the real-time thermal diffusion state of the weed tissue.

[0055] In the embodiment of the present application, step 51 constructs an initial polygonal boundary network based on a high-thermal-response focus, so that the boundary can accurately surround the core thermal response area, ensuring that the energy action range is focused on the key inactivation area of the weed, and avoiding that the initial boundary excessively covers the non-target area; the boundary smoothing processing of step 52 eliminates sharp corners, so that the initial boundary is more consistent with the continuous distribution characteristics of the weed tissue, and reduces the problem of uneven energy distribution caused by irregular boundaries; step 53 calculates the thermal diffusion rate field, which can capture the spatial gradient of heat propagation in real time, and provides a quantitative basis for reflecting the actual trend of thermal diffusion for the dynamic adjustment of the boundary, ensuring that the adjustment direction conforms to the real thermal behavior of the weed tissue; step 54 predicts the anisotropic thermal diffusion trend, which can identify the thermal diffusion demand of the boundary in each direction and provide clear direction guidance for differentiated adjustment, avoiding energy waste or insufficient coverage caused by indiscriminate adjustment; step 55 dynamically adjusts the boundary form according to the prediction trend, so that the action range is adapted to the thermal diffusion in real time, expands in the high-thermal-diffusion area to cover the newly heated area, and contracts in the low-thermal-diffusion area to reduce invalid energy consumption, finally realizing the dynamic optimization of the energy action range, which not only ensures the global inactivation of the weed, but also significantly improves the energy utilization efficiency.

[0056] In a preferred embodiment of the present application, step 6 divides the action range boundary into several independent energy regulation areas according to the thermodynamic gradient characteristics, including: Step 61, based on the action range boundary form data output by step 55 after dynamic adjustment at the current time, extracts the closed area surrounded by the boundary as the basic regulation range; Step 62, based on the real-time updated spatiotemporal temperature field data output by step 35, calculates the temperature change rate gradient value and the thermal cumulative dose value of each spatial position point in the basic regulation range; Step 63, according to the temperature change rate gradient value and the thermal cumulative dose value calculated in step 62, a thermodynamic gradient distribution map covering the basic regulation range is generated; Step 64, the thermodynamic gradient distribution map is processed by isogradient line division to form sub-regions with similar thermodynamic gradient characteristics; Step 65, each continuous sub-region with similar thermodynamic gradient characteristics obtained by step 64 is defined as an independent energy regulation area, and the spatial range identification of all independent energy regulation areas is output.

[0057] In the embodiment of the present application, the specific implementation process of step 61 is as follows: From the dynamic adjustment range boundary form data output in step 55, the structured data (usually coordinate array, format like [(X1, Y1), (X2, Y2),..., (Xn, Yn)]) storing boundary coordinates is read. Each coordinate point in the array is extracted one by one, and its specific values of X axis and Y axis (such as units of centimeters, accurate to 0.01 cm) are recorded to ensure that there is no coordinate point missing or repeated; by calculating the direction of the line connecting adjacent coordinate points (such as comparing the offset direction of the next point relative to the previous point), it is determined whether the coordinate points are arranged in clockwise or counterclockwise order. For example, if the line from (X1, Y1) to (X2, Y2) to (X3, Y3) always deflects to the same side (such as all to the left), it is determined that the arrangement is in order; by comparing the values of the first coordinate point (X1, Y1) and the last coordinate point (Xn, Yn) in the array, if the X difference of the two is ≤0.01 cm and the Y difference is ≤0.01 cm (the error threshold is set to allow slight calculation error), it is determined that the boundary is continuous; if the difference exceeds the threshold, the data generation process in step 55 needs to be traced back to correct the boundary coordinates to eliminate gaps and ensure that the polygon is not broken.

[0058] The ray method is used as the core algorithm for spatial region clipping, and the specific operation is as follows: According to the extreme values of the boundary coordinates (such as the minimum X value Xmin, the maximum X value Xmax, the minimum Y value Ymin, and the maximum Y value Ymax), the rectangular boundary of the region to be verified is determined (X∈[Xmin-0.5cm, Xmax+0.5cm], Y∈[Ymin-0.5cm, Ymax+0.5cm]) to ensure that the boundary and the surrounding area are covered and the points near the boundary are not missed; within the above rectangular range, evenly distributed verification points are generated according to the preset grid spacing (such as 0.05cm×0.05cm, taking into account accuracy and efficiency), and the coordinates of each point are where, Starting from Xmin-0.5cm, incrementing by 0.05cm to Xmax+0.5cm; Similarly, for each grid check point, perform the "point-in-polygon" judgment: draw a virtual ray from the point to the horizontal right side (X-axis positive direction), and count the number of intersection points of the ray and the polygon boundary. If the number of intersection points is odd, it is determined that the point is inside the polygon (marked as "within the basic control range"); if it is even or 0, it is determined to be outside (marked as "outside the range"). If the ray exactly passes through the boundary coordinate point or coincides with the boundary edge, the direction of the ray needs to be adjusted (such as shifting 0.001 cm to the Y-axis positive direction) to recalculate to avoid misjudgment. After completing the check of all grid points, all points marked as "within the basic control range" are summarized to form a complete spatial boundary of the basic control range, ensuring that only points within this range are executed in subsequent steps (such as temperature parameter calculation in step 62), and completely excluding non-target areas such as soil and crops outside the boundary.

[0059] The specific implementation process of the above step 62 is as follows: According to the spatial boundary of the basic control range (defined in step 61), set the grid parameters: the grid spacing is 0.1 cm x 0.1 cm, that is, along the X-axis and Y-axis directions, every 0.1 cm is divided into a grid point, ensuring that all grid points fall within the "basic control range". Through coordinate mapping, each grid point is assigned a unique coordinate identifier (such as (X1, Y1), (X1, Y1+0.1 cm), etc.); From the real-time space-time temperature field data output in step 35, extract the temperature change rate of each grid point at the current time (i.e. the "unit time temperature change" calculated in step 41, unit: ℃ / ms). When extracting, the data correspondence needs to be strictly checked: through the spatial coordinate matching of grid point coordinates and temperature field data, ensure that the temperature change rate of each grid point is accurate (such as excluding mismatch caused by data transmission delay). If a grid point has no corresponding temperature data (such as edge data missing), use the average value of its nearest 3 valid grid points to fill in to avoid data gaps; For each grid point, select adjacent points according to the "8-neighborhood rule": i.e. the upper (X, Y+0.1 cm), lower (X, Y-0.1 cm), left (X-0.1 cm, Y), right (X+0.1 cm, Y), upper left (X-0.1 cm, Y+0.1 cm), upper right (X+0.1 cm, Y+0.1 cm), lower left (X-0.1 cm, Y-0.1 cm), and lower right (X+0.1 cm, Y-0.1 cm) adjacent grid points in 8 directions; For boundary grid points (such as points on the edge of the basic control range), if there is no adjacent grid point in a certain direction (such as the leftmost point has no left neighbor), exclude that direction and only keep the existing adjacent points (such as the leftmost point retains 7 or fewer adjacent points), and record the missing direction to avoid invalid calculation.

[0060] Calculate the temperature rate difference between the current point and each adjacent point: subtract the temperature rate of the adjacent point from the temperature rate of the current point, and take the absolute value (e.g., if the current point rate is 0.8°C / ms and the right point is 0.6°C / ms, the difference is |0.8-0.6| = 0.2°C / ms). Repeat this operation for all valid adjacent points to obtain a set of difference data (e.g., 8, 7, etc.). Summarize the difference data of all adjacent points and calculate the comprehensive gradient value using the "arithmetic mean method": add all the differences and divide by the number of valid adjacent points (e.g., divide by 8 for 8 adjacent points or by 7 for 7 adjacent points) to obtain the "temperature rate gradient value" of the grid point. For example, the differences between a certain point and 8 adjacent points are 0.2, 0.3, 0.1, 0.2, 0.4, 0.3, 0.2, and 0.1°C / ms, the total is 1.8°C / ms, and the gradient value is 1.8 ÷ 8 = 0.225°C / ms. The higher the gradient value, the more significant the difference between the point and the surrounding area (e.g., the difference between the core area and the edge area of the weed).

[0061] Check the continuity of the time series during extraction: if there are missing time points (e.g., no temperature data for a 0.01ms interval), use linear interpolation to supplement (e.g., estimate the temperature of tk+1 based on the temperatures of tk and tk+2), ensuring that the sequence is continuous. Set the time interval Δt = 0.01ms (match the temperature sampling frequency to ensure that every temperature change detail is covered). Count the total number of intervals n in the historical temperature sequence: from t0 to tn, there are n intervals (n = (tn-t0) / 0.01ms).

[0062] The temperature values T(ti) (i=0 to n-1) of each time interval are added to obtain the total temperature sum T = T(t0) + T(t1) +... + T(tn-1); and the total temperature sum is multiplied by the time interval At, i.e. the heat cumulative dose = T x 0.01 ms. For example, if the total temperature value of 10 intervals is 50°C, the dose = 50 x 0.01 = 0.5°C ms, which reflects the total heat intensity received by the point from the current time to the current time; if the heat cumulative dose value is negative (not possible) or far exceeds the average level of similar grid points (such as more than 3 times the average), the temperature sequence extraction process is traced back to check if there is a data error (such as abnormal temperature value) and recalculate to ensure that the dose value accurately reflects the actual heat accumulation; all grid points in the basic control range are traversed in the order of "row by column": starting from the leftmost X smallest column, each grid point is processed in turn along the Y axis direction, and after completing a column, it is moved to the next column until all points are processed; for each grid point, the calculated "temperature change rate gradient value" and "heat cumulative dose value" are bound with the coordinates of the point and stored as structured data (such as each entry containing coordinates (X, Y), gradient value, dose value) to ensure that the corresponding parameters can be directly called in the subsequent steps (such as step 63 to generate a thermodynamic gradient distribution map). After the traversal is completed, the dataset containing all grid point parameters is output as the input of step 63.

[0063] The specific implementation process of the above step 63 is as follows: A two-dimensional coordinate system (consistent with the spatial coordinates of the basic control range) is established, and the "temperature change rate gradient value" and "heat cumulative dose value" of each spatial point calculated in step 62 are mapped to the corresponding positions of the coordinate system; data visualization mapping rules are used: for example, the temperature change rate gradient value is represented by color depth (dark color represents high gradient, light color represents low gradient), and the heat cumulative dose value is represented by transparency (high transparency represents low dose, low transparency represents high dose). By superimposing the characteristics of the two parameters, a thermodynamic gradient distribution map covering the entire basic control range is generated, and the visual features (color + transparency) of each position in the map intuitively reflect the thermodynamic gradient comprehensive characteristics (heat response difference degree and total heat accumulation state) of the point. The distribution map is data checked to ensure that the gradient features of each spatial point are accurately mapped without data loss or mismatch (such as whether the gradient value of the edge point is reasonable).

[0064] The specific implementation process of the above step 64 is as follows: Based on the thermodynamic gradient distribution map generated in step 63, the temperature change rate gradient value and the thermal cumulative dose value of all spatial points are extracted, and the value range of the two parameters is determined (such as gradient value 0.1-1.0℃ / ms, dose value 5-50℃·ms); according to the distribution characteristics of the gradient value and the dose value, several continuous intervals are divided (such as 0.2℃ / ms for each interval of the gradient value, and 10℃·ms for each interval of the dose value), to ensure that the parameter values in each interval are small (i.e. "similar thermodynamic gradient characteristics"); starting from any unmarked point, check whether the gradient value and the dose value of its adjacent points fall into the same interval (such as the current point gradient 0.3-0.5℃ / ms, dose 10-20℃·ms, the adjacent points need to meet the same interval), if it meets the requirements, the adjacent points are included in the same sub-region; repeat the expansion until there is no adjacent point that meets the requirements, and form a sub-region. Repeat this process for all unmarked points, and finally obtain multiple continuous sub-regions that are independent of each other and have similar internal parameter characteristics.

[0065] The specific implementation process of step 65 is as follows: For each sub-region divided in step 64, the spatial boundary of the sub-region is determined by the boundary extraction algorithm (such as the coordinate point set of the outer periphery of the sub-region), and the continuity of the region is verified (i.e. no blank or fracture in the interior, all points are connected); a unique identifier is assigned to each continuous sub-region (such as "control region 1" "control region 2"), and the spatial range characteristics are recorded: for example, the rectangular boundary range is represented by the minimum / maximum X coordinate and Y coordinate of all points in the sub-region, or the boundary coordinate sequence of the sub-region is directly recorded; the identifier is bound with the spatial range characteristics to form an "independent energy control region list", each entry of which contains the control region ID, the boundary coordinate range, the core thermodynamic gradient characteristics (such as the average gradient value, the average dose value), to ensure that the spatial position and characteristics of each control region can be accurately located by the list in the subsequent steps, and finally the list is output as the result of step 6.

[0066] The step 61 extracts the closed basic regulation range, ensures that the energy regulation is only focused on the effective action area after dynamic adjustment, avoids invalid coverage of non-target areas outside the boundary, and locks the regulation object in space. The step 62 calculates the temperature change rate gradient value and the thermal cumulative dose value of each point, quantifies the thermal dynamic differences (such as thermal response speed and total thermal cumulative amount) of the weed tissue into analyzable parameters, and provides fine thermodynamic characteristic data support for subsequent zoning. The step 63 generates a thermodynamic gradient distribution map, intuitively presents the thermal characteristic spatial distribution in the basic regulation range, and makes the thermodynamic differences of different regions obvious at a glance, thereby providing a visual reference for zoning division. The step 64 forms a similar characteristic sub-region through equi-gradient line division, aggregates the regions with similar thermodynamic characteristics, and ensures that the weed thermal response law in each sub-region is consistent. The step 65 defines an independent energy regulation area and outputs an identifier, so that subsequent energy adjustment can be executed according to the regional differences, the energy of the region with insufficient thermal response is enhanced, the output of the region with sufficient thermal accumulation is reduced, the energy is supplied according to the need, the accuracy and efficiency of energy regulation are significantly improved, and the problems of energy waste or incomplete local inactivation are avoided.

[0067] In a preferred embodiment of the present application, the step 7 comprises: The step 71 obtains the thermal cumulative dose value of all points in each independent energy regulation area at the current moment, calculates the average value of the thermal cumulative dose of all points in the regulation area as the current energy cumulative state characteristic value of the regulation area, obtains the temperature value of all points in the regulation area at the current moment, and calculates the proportion of pixel points whose temperature value reaches the critical temperature threshold of tissue inactivation in the regulation area as the current thermal damage degree characteristic value of the regulation area. The step 72 compares the current energy cumulative state characteristic value with the preset target cumulative energy value of the regulation area, and calculates the energy cumulative difference value. The step 73 compares the current thermal damage degree characteristic value with the preset target thermal damage degree threshold of the regulation area, and calculates the thermal damage degree difference value. The step 74 calculates the energy compensation demand intensity of the regulation area according to the energy cumulative difference value and the thermal damage degree difference value. The step 75 normalizes the energy compensation demand intensity to generate the zoning energy compensation coefficient of the regulation area, and generates the respective zoning energy compensation coefficient for all independent energy regulation areas.

[0068] In the embodiment of the present application, the specific implementation process of the step 71 is as follows: The spatial range of each regulation zone is accurately positioned according to the spatial range identification of the independent energy regulation zone output in step 65 (such as the boundary coordinates of each regulation zone). For each regulation zone, the "thermal cumulative dose value" and "current time temperature value" of all spatial points (grid points) within the range are extracted from the structured data generated in step 62, excluding points outside the regulation zone and invalid data (such as points with abnormal temperature values, which are determined to be invalid by a preset threshold, such as temperature <0°C or >100°C, and are replaced by the mean value of adjacent valid points); the thermal cumulative dose values extracted are summarized: the total number of valid points in the regulation zone is counted (denoted as N), and the sum of the thermal cumulative dose values of all valid points is added to obtain the total sum (denoted as S). The average value is calculated: the current energy accumulation state representation value = S ÷ N. For example, a regulation zone has 100 valid points, and the total dose sum is 500°C·ms, then the average value = 500 ÷ 100 = 5°C·ms, which reflects the overall heat accumulation level of the region; a preset tissue inactivation threshold temperature (based on experimental data, such as the critical temperature for complete inactivation of weed cells being 60°C). The temperature values in the regulation zone are counted: the number of valid points with temperature values ≥ 60°C is counted (denoted as M), and the proportion is calculated: the current heat damage degree representation value = M ÷ N × 100%. For example, among the 100 valid points, 65 reach 60°C, then the proportion = 65 ÷ 100 × 100% = 65%, which reflects the proportion of weed tissue in the region that has been heat damaged.

[0069] The specific implementation process of the above step 72 is as follows: A preset target cumulative energy value for each regulation zone is set based on the thermodynamic gradient characteristics of the regulation zone (such as the gradient distribution map in step 63) and the type of weeds (the type of weeds associated through previous positioning data), and a differentiated target value is set. For example, the target value of a regulation zone with slower heat response (low gradient value) is set to 8°C·ms, and the target value of a region with faster heat response is set to 6°C·ms, to ensure that the target matches the characteristics of the region; the "current energy accumulation state representation value" obtained in step 71 is compared with the preset target cumulative energy value, and the difference is calculated: energy accumulation difference = target cumulative energy value - current energy accumulation average value. For example, the target value is 8°C·ms, and the current average value is 5°C·ms, then the difference = 8-5 = 3°C·ms (a positive value indicates that the energy accumulation is insufficient and needs to be supplemented; a negative value indicates that it is excessive).

[0070] The specific implementation process of the above step 73 is as follows: Pre-set the target heat damage threshold of each control area (based on the herbicide requirement, such as 90% of the weed tissue reaching the inactivation temperature), the threshold can be adjusted according to the importance of the control area, such as the threshold of the core high-temperature response area is set to 95%, and the threshold of the edge area is set to 85%; compare the "current heat damage degree representation value" (proportion) obtained in step 71 with the target threshold, and calculate the difference: heat damage degree difference = target threshold - current heat damage proportion. For example, the target threshold is 90%, and the current proportion is 65%, then the difference = 90%-65%=25% (a positive value indicates that the damage is insufficient and needs to be strengthened; a negative value indicates that the damage is excessive).

[0071] The specific implementation process of the above step 74 is as follows: According to the influence weight of energy accumulation and heat damage on the herbicidal effect, pre-set the weight coefficients of the two (such as the energy accumulation difference weight 0.6 and the heat damage degree difference weight 0.4, which are determined based on the contribution of the two to the inactivation effect in the experiment); multiply the energy accumulation difference of step 72 and the heat damage degree difference of step 73 by the corresponding weight, and then sum to obtain the demand intensity: energy compensation demand intensity = (energy accumulation difference x 0.6) + (heat damage degree difference x 0.4). For example, the energy difference is 3℃·ms, and the heat damage difference is 25%, then the demand intensity = 3x0.6+25% x 0.4 = 1.8+0.1 = 1.9 (the larger the value, the stronger the compensation demand). If a difference is negative (such as energy excess), then this part is calculated as 0 to avoid reverse compensation.

[0072] The specific implementation process of the above step 75 is as follows: Collect the energy compensation demand intensity values of all independent energy control areas, and find the maximum value (denoted as Max) and the minimum value (denoted as Min). If all demand intensities are 0, then the compensation coefficients are all set to 0; otherwise, normalize the demand intensity of each control area: partition energy compensation coefficient = (current demand intensity-Min) ÷ (Max-Min). For example, the demand intensity of a certain control area is 1.9, the Max of all areas is 2.0, and the Min is 0.5, then the coefficient = (1.9-0.5) ÷ (2.0-0.5) = 1.4 ÷ 1.5 ≈ 0.93 (the coefficient range is 0-1, and the closer to 1 indicates that the compensation demand is stronger); bind the normalized coefficient of each control area with its spatial range identifier to generate a "control area ID-compensation coefficient" corresponding table, which ensures that the subsequent energy adjustment can accurately match the compensation intensity according to the area, and output the table as the input basis of step 8. Through the above process, step 7 realizes the quantitative evaluation of the energy state of each control area and the accurate calculation of the compensation demand.

[0073] In a preferred embodiment of the application, the step 8 comprises: Step 81, based on the partition energy compensation coefficient of each independent energy regulation zone generated in step 75, a mapping relationship table of regulation zone space identification and compensation coefficient is established; Step 82, according to the mapping relationship table, the required high-frequency current output increment proportion of each independent energy regulation zone is analyzed, and the corresponding electrode control instruction set is generated; Step 83, the electrode control instruction set is executed, and the high-frequency current generator is driven to implement differentiated current output to each independent energy regulation zone divided in step 6; Step 84, in the process of differentiated current output, the operations of steps 3 to 7 are repeatedly executed synchronously to calculate the updated partition energy compensation coefficient; Step 85, it is judged whether the updated thermal damage degree representation value of each regulation zone in step 71 reaches the preset target thermal damage degree threshold, and whether the updated energy accumulation state representation value of each regulation zone in step 71 reaches the corresponding target cumulative energy value; Step 86, if the result of step 85 is not satisfied, return to step 83 to continue to execute differentiated adjustment; Step 87, if the result of step 85 is satisfied, it is determined that the global inactivation condition is reached, and the high-frequency current generator is controlled to terminate energy output.

[0074] In the embodiment of the present application, the specific implementation process of step 81 is as follows: From the output result of step 75, the "space range identification" (such as regulation zone ID, boundary coordinate range) and the corresponding "partition energy compensation coefficient" (normalized 0-1 value) of all independent energy regulation zones are extracted. These two types of data are arranged according to the "one-to-one correspondence" principle, for example, "regulation zone 1 (ID: 001) corresponds to compensation coefficient 0.93", "regulation zone 2 (ID: 002) corresponds to compensation coefficient 0.45", etc. A structured table form is used, and the table fields include "regulation zone ID", "spatial boundary coordinates (Xmin, Xmax, Ymin, Ymax)", "partition energy compensation coefficient", and "data generation timestamp". After entering the data, the integrity of the table is checked: check whether there is a missing regulation zone (compared with the total number of regulation zones output in step 65), whether the compensation coefficient is out of the range of 0-1 (if so, go back to step 75 to correct), and ensure that each regulation zone has a unique compensation coefficient, providing accurate mapping basis for subsequent instruction analysis.

[0075] The specific implementation process of step 82 is as follows: The conversion rule of the preset compensation coefficient and the increment ratio is as follows: when the compensation coefficient is 0, the increment ratio is 0% (no additional output); when the compensation coefficient is 1, the increment ratio is 50% (the maximum allowed increment to avoid excessive energy output); and the intermediate coefficient is converted in a linear relationship, that is, "the increment ratio = the compensation coefficient x 50%". For example, the compensation coefficient 0.93 corresponds to the increment ratio = 0.93 x 50% ≈ 46.5%, and the compensation coefficient 0.45 corresponds to the increment ratio = 0.45 x 50% = 22.5%; for each regulation area, according to the compensation coefficient in the mapping relationship table, the required high-frequency current output increment ratio (the increment object includes the voltage amplitude or the action time, and the voltage amplitude is preferentially adjusted) is calculated according to the above rule. For example, if the voltage amplitude is initially 1400 volts and the increment is 46.5%, the voltage amplitude is adjusted to 1400 x (1 + 46.5%) ≈ 2051 volts.

[0076] The electrode control instruction set needs to include the following information: regulation area ID, corresponding electrode array number (the electrode array is bound to the regulation area according to spatial partitioning), target voltage amplitude (initial value + increment), target action time (if adjustment is needed), and instruction execution timestamp. For example, "regulation area ID: 001, electrode number: E01, target voltage: 2051 volts, action time: 1.3 seconds (maintain the initial value), and execution time: t + 0.05 ms".

[0077] The legality of the instruction set is verified: the target voltage is checked to be within the device output range (such as the 1600-2000 volt level), the electrode number is matched with the regulation area, and it is ensured that the instruction can be recognized and executed by the high-frequency current generator, and finally the formatted instruction set (such as in JSON format) is output.

[0078] The specific implementation process of the above step 83 is as follows: The electrode control instruction set generated in step 82 is sent to the control system of the high-frequency current generator through the data interface, and the control system parses the regulation area ID, electrode number and target parameters in the instruction; for each electrode array corresponding to the regulation area, the control system adjusts the output parameters of its driving module to make the voltage amplitude (or action time) of the high-frequency current increase by the calculated increment ratio. For example, the electrode E01 corresponding to the regulation area 001 has an original output voltage of 1400 volts, which is adjusted to 2051 volts by an increment of 46.5%, while the frequency remains unchanged at 80 kHz; during execution, the actual output parameters (voltage, current, time) are monitored in real time by the built-in sensor of the generator, and compared with the target values of the instruction. If the deviation exceeds 5% (the preset allowed error), the fine adjustment (such as increasing the driving signal when the voltage is insufficient) is triggered immediately to ensure that the current output of each regulation area accurately meets the increment demand.

[0079] The specific implementation process of the above step 84 is as follows: After the differentiated current output starts, a synchronous update cycle is set (e.g., once every 0.1 seconds, adjusted according to the thermal response speed). Within each cycle, the complete process of steps 3 to 7 is re-executed to generate new partition energy compensation coefficients; after each update, the new coefficients are compared with the previous cycle coefficients, and if the change exceeds 10% (preset sensitivity threshold), it is marked as "need to focus on the regulation area" to provide priority reference for subsequent adjustment.

[0080] The specific implementation process of the above step 85 is as follows: The updated data of each regulation area in step 84 is collected and two conditions are checked one by one to check whether the "current thermal damage degree representation value" (e.g., proportion) of each regulation area is ≥ the preset target thermal damage degree threshold (e.g., 90%). For example, the current proportion of regulation area 001 is 92% ≥ 90%, which is determined to meet the condition; the current proportion of regulation area 002 is 88% < 90%, which is determined to not meet the condition; the "current energy accumulation state representation value" (average value) of each regulation area is checked to see whether it is ≥ the preset target cumulative energy value (e.g., 8℃·ms). For example, the average value of regulation area 001 is 8.2℃·ms ≥ 8℃·ms, which meets the condition; the average value of regulation area 002 is 7.5℃·ms < 8℃·ms, which does not meet the condition; only when both conditions of all regulation areas meet (i.e., 100% of the regulation areas pass the check), it is determined that the global inactivation condition is met; if there is any regulation area that does not meet the condition, it is determined that the condition is not met.

[0081] The specific implementation process of the above step 86 is as follows: If step 85 determines that the condition is not met, the system automatically backtracks to step 83: based on the updated partition energy compensation coefficients in step 84, the electrode control instruction set is re-generated (the analysis process of step 82 is repeated, and the incremental proportion is adjusted according to the new coefficients), and the high-frequency current generator is driven to continue implementing differentiated current output on the regulation areas that do not meet the condition (the energy increment of the areas that do not meet the condition is highlighted); at the same time, the identification and non-meeting items (e.g., "regulation area 002 thermal damage is insufficient" and "regulation area 003 energy accumulation is insufficient") of the regulation areas that do not meet the condition are recorded, and the compensation effort on the corresponding areas in the new instruction set is increased (e.g., the incremental proportion upper limit is appropriately increased to 60%) to accelerate the process of meeting the condition.

[0082] The specific implementation process of the above step 87 is as follows: If step 85 determines that the condition is met (all control regions meet the standard), the system sends a "stop output command" to the high-frequency current generator, which includes the stop time, the gradual rate of voltage amplitude reduction to 0 volts (such as 500 volts / ms, to avoid current shock to the equipment); after receiving the command, the generator gradually reduces the output voltage to 0 volts at the gradual rate, and cuts off the current supply to the electrode array. Subsequently, the system records the complete parameters of this time of weed removal (energy output of each control region, total time consumption, standard time, etc.), generates a weed removal completion log, and enters a standby state, waiting for the next round of weed positioning signal to trigger a new operation process; through the above process, step 8 realizes dynamic closed-loop control of energy output, ensuring that weeds are inactivated under the premise of precise energy consumption, which not only improves the reliability of weed removal, but also avoids energy waste.

[0083] As Figure 2 shown, the embodiment of the present application also provides an energy output dynamic control system of a high-frequency electric shock weed removal device, comprising: An acquisition module for identifying the above-ground morphological characteristics of the target weeds through a visual sensor, and generating positioning data containing plant size and spatial distribution; A calculation module for calculating the voltage amplitude, frequency and action time parameters of the initial high-frequency current based on the plant size in the positioning data according to a preset mapping relationship; executing the initial energy parameters and applying high-frequency current to the weeds, while collecting the thermodynamic response signals of the weed tissues in real time; A calibration module for calibrating a plurality of high-thermal-response focal points in the energy action area according to the temperature rise rate distribution in the thermodynamic response signals; connecting the high-thermal-response focal points to construct a dynamic action range boundary, and adjusting the boundary shape according to the real-time thermal diffusion trend; A division module for dividing the action range boundary into a plurality of independent energy control regions according to the thermodynamic gradient characteristics; generating a partition energy compensation coefficient based on the thermal damage degree and energy accumulation state of each control region; and differentially adjusting the high-frequency current output according to the partition energy compensation coefficient until the thermodynamic response signals meet the global inactivation condition.

[0084] The above is the preferred embodiment of the present application. It should be noted that for ordinary skilled persons in the technical field, without departing from the principles of the present application, a number of improvements and refinements can be made, which should also be considered within the scope of protection of the present application.

Claims

1. A method for dynamic control of energy output of a high-frequency electric shock weeding device, characterized in that, The method comprises: Step 1, identifying the above-ground morphological characteristics of the target weed through a visual sensor, and generating positioning data containing plant size and spatial distribution; Step 2, based on the plant size in the positioning data, calculating the voltage amplitude, frequency and action time parameters of the initial high-frequency current according to a preset mapping relationship; Step 3, executing the initial energy parameters and applying high-frequency current to the weeds, while real-time collecting the thermodynamic response signals of the weed tissues; Step 4, according to the temperature rise rate distribution in the thermodynamic response signals, calibrating multiple high-heat response foci in the energy action area; Step 5, connecting the high-heat response foci to construct a dynamic action range boundary, and adjusting the boundary shape according to the real-time heat diffusion trend; Step 6, dividing the action range boundary into several independent energy regulation zones according to the thermodynamic gradient characteristics; Step 7, based on the thermal damage degree and energy accumulation state of each regulation zone, generating a partition energy compensation coefficient; Step 8, differentially adjusting the high-frequency current output according to the partition energy compensation coefficient difference until the thermodynamic response signal meets the global inactivation condition.

2. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 1, characterized by, The step 1 comprises: Step 11, synchronously capturing the left and right perspective images of the target area through the calibrated binocular visual sensor; Step 12, performing vegetation spectral response enhancement processing on the left and right perspective images respectively, segmenting out the weed pixel area and generating a binary contour mask; Step 13, extracting feature matching point sets based on the binary contour mask, and generating a depth mapping graph through binocular disparity calculation; Step 14, converting the depth mapping graph into a three-dimensional space point cloud model, and fitting a ground reference plane; Step 15, calculating the vertical distance from the highest point of each connected domain in the point cloud model to the ground reference plane as the plant height, and the minimum circumscribed circle diameter of the point cloud projection on the horizontal plane as the crown diameter; Step 16, taking the bottom centroid coordinates of the point cloud connected domain as the plant position reference, and outputting the structured positioning data containing three-dimensional coordinates, plant height and crown diameter.

3. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 2, characterized in that, The step 2 comprises: Step 21, according to the plant height value in the positioning data, referring to the mapping table of preset plant height interval and voltage amplitude reference level to determine the corresponding voltage amplitude reference level; wherein the mapping table is set to correspond to different voltage reference levels when the plant height value falls into different threshold ranges, and the higher the plant height value, the higher the corresponding voltage reference level; Step 22, according to the crown diameter value in the positioning data, calculating the initial action time according to the preset positive correlation mapping relationship between the crown diameter value and the initial action time; wherein the positive correlation mapping relationship is set to the larger the crown diameter value, the longer the calculated initial action time; Step 23, setting a fixed high-frequency frequency parameter value, which is configured to be higher than a preset human safety frequency threshold, while being lower than a preset electromagnetic interference critical frequency threshold; Step 24, combine the voltage amplitude reference level determined in step 21 as the initial voltage amplitude, the initial action duration calculated in step 22 as the initial action duration, and the fixed high-frequency frequency parameter value set in step 23 to generate the initial high-frequency current voltage amplitude, frequency, and action duration parameter combination.

4. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 3, characterized in that, The step 3 includes: Step 31, based on the initial high-frequency current voltage amplitude, frequency, and action duration parameter combination generated in step 24, control the high-frequency current generator to output a high-frequency current pulse sequence with corresponding voltage amplitude, frequency, and planned action duration; Step 32, apply the high-frequency current pulse sequence to the target weed plants through the contact electrode array, and the spatial arrangement of the electrode array covers the weed spatial distribution range indicated by the positioning data generated in step 1; Step 33, during the entire process of applying the high-frequency current pulse sequence, use the infrared thermal imager arranged in the gap or adjacent position of the electrode array to continuously capture the infrared thermal radiation image sequence of the target weed plants and the surrounding action area at a sampling rate higher than the high-frequency current pulse frequency; Step 34, perform time series analysis on the infrared thermal radiation image sequence to extract the continuous temperature value of each spatial pixel point during the current action, and generate the space-time temperature field data reflecting the real-time temperature change of the weed tissue internal; Step 35, output the space-time temperature field data as a signal representing the thermodynamic response of the weed tissue.

5. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 4, characterized in that, Step 4, according to the temperature rise rate distribution in the thermodynamic response signal, label multiple high-heat response foci in the energy action area, including: Step 41, based on the space-time temperature field data output in step 35, calculate the temperature change per unit time of each spatial pixel point in the energy action area within a preset time window to generate corresponding temperature rise rate field data; Step 42, perform spatial domain analysis on the temperature rise rate field data to identify all pixel points with a temperature rise rate value exceeding a preset first rate threshold to form a candidate high-heat response point set; Step 43, perform spatial clustering processing on the candidate high-heat response point set to aggregate candidate points meeting a preset spatial proximity condition into several independent high-heat response clusters; Step 44, calculate the average temperature rise rate of all points in each high-heat response cluster, and select the cluster with an average value exceeding a preset second rate threshold as an effective high-heat response cluster; Step 45, determine the spatial center coordinates of each effective high-heat response cluster, and label the center coordinates as a high-heat response focus in the energy action area, and output the coordinate set of all labeled high-heat response foci.

6. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 5, characterized in that, The step 5 includes: Step 51, based on the coordinate set of all labeled high-heat response foci output in step 45, perform spatial topology connection operation to connect adjacent foci meeting a preset distance threshold condition to form an initial polygon boundary network; Step 52, perform boundary smoothing processing on the initial polygon boundary network to eliminate local sharp corners and generate a continuous closed initial action range boundary contour; Step 53, based on the real-time updated spatiotemporal temperature field data output in step 35, calculate the heat diffusion rate field data of the current time in the energy action area; the heat diffusion rate field data reflects the spatial gradient characteristics of heat propagation from high-temperature area to low-temperature area; Step 54, according to the heat diffusion rate field data, predict the anisotropic heat diffusion trend within a preset distance from the outer edge of the initial action range boundary contour; wherein the area with high heat diffusion rate is predicted as the boundary expansion direction, and the area with low heat diffusion rate is predicted as the boundary maintenance or contraction direction; Step 55, according to the anisotropic heat diffusion trend predicted in step 54, perform real-time deformation adjustment on the initial action range boundary contour: adjust the boundary to extend outward by a preset deformation amount in the predicted expansion direction, or keep or contract the boundary inward in the predicted maintenance or contraction direction; generate a dynamically updated action range boundary shape, and output the action range boundary shape data after dynamic adjustment at the current time.

7. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 6, characterized in that, Step 6, divide the action range boundary into several independent energy regulation areas according to the thermodynamic gradient characteristics, including: Step 61, based on the current time dynamic adjustment action range boundary shape data output in step 55, extract the closed area surrounded by the boundary as the basic regulation range; Step 62, based on the real-time updated spatiotemporal temperature field data output in step 35, calculate the temperature change rate gradient value and heat cumulative dose value of each spatial position point in the basic regulation range; Step 63, according to the temperature change rate gradient value and heat cumulative dose value of each point calculated in step 62, generate a thermodynamic gradient distribution map covering the basic regulation range; Step 64, perform equal gradient line division processing on the thermodynamic gradient distribution map to form sub-regions with similar thermodynamic gradient characteristics; Step 65, define each continuous sub-region with similar thermodynamic gradient characteristics obtained in step 64 as an independent energy regulation area, and output the spatial range identifier of all independent energy regulation areas.

8. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 7, characterized in that, The step 7 includes: Step 71, obtain the heat cumulative dose value of all points in each independent energy regulation area at the current time; calculate the average value of the heat cumulative dose of all points in the regulation area as the current energy accumulation state characteristic value of the regulation area; obtain the temperature value of all points in the regulation area at the current time; and calculate the proportion of pixel points whose temperature value reaches the critical temperature threshold of tissue inactivation in the regulation area as the current heat damage degree characteristic value of the regulation area; Step 72, compare the current energy accumulation state characteristic value with the preset target cumulative energy value of the regulation area to calculate the energy accumulation difference value; Step 73, compare the current heat damage degree characteristic value with the preset target heat damage degree threshold of the regulation area to calculate the heat damage degree difference value; Step 74, calculate the energy compensation demand intensity of the regulation area according to the energy accumulation difference value and the heat damage degree difference value; Step 75, normalize the energy compensation demand intensity to generate the partition energy compensation coefficient of the regulation area; generate the partition energy compensation coefficient of each independent energy regulation area.

9. The energy output dynamic control method of the high-frequency electric shock weeding apparatus according to claim 8, characterized in that, The step 8 includes: Step 81, based on the partition energy compensation coefficient of each independent energy regulation zone generated in step 75, a mapping relationship table of regulation zone space identifier and compensation coefficient is established; Step 82, according to the mapping relationship table, the required high-frequency current output increment proportion of each independent energy regulation zone is analyzed, and the corresponding electrode control instruction set is generated; Step 83, execute the electrode control instruction set to drive the high-frequency current generator to implement differentiated current output to each independent energy regulation zone divided in step 6; Step 84, during the process of differentiated current output, the operations of steps 3 to 7 are repeatedly executed synchronously to calculate the updated partition energy compensation coefficient; Step 85, judge whether the updated thermal damage degree characteristic value of each regulation zone in step 71 reaches the preset target thermal damage degree threshold, and whether the updated energy accumulation state characteristic value of each regulation zone in step 71 reaches the corresponding target cumulative energy value; Step 86, if the judgment result of step 85 is not satisfied, return to step 83 to continue to execute the differentiated adjustment; Step 87, if the judgment result of step 85 is satisfied, it is judged that the global inactivation condition is reached, and the high-frequency current generator is controlled to terminate energy output.

10. A dynamic energy output control system for a high frequency electrical shock weed control apparatus, the system implementing the method of any one of claims 1 to 9, characterised in that, Comprise: The acquisition module is used for identifying the aboveground morphological characteristics of the target weeds through the visual sensor, and generating positioning data containing plant size and spatial distribution; The calculation module is used for calculating the voltage amplitude, frequency and action time parameters of the initial high-frequency current according to the preset mapping relationship based on the plant size in the positioning data; Execute the initial energy parameters and apply high-frequency current to weeds, while collecting the thermodynamic response signals of weed tissues in real time; The calibration module is used for calibrating a plurality of high-heat response focal points in the energy action area according to the temperature rise rate distribution in the thermodynamic response signal; Connect the high-heat response focal points to construct a dynamic action range boundary, and adjust the boundary shape according to the real-time heat diffusion trend; The division module is used for dividing the action range boundary into a plurality of independent energy regulation zones according to the thermodynamic gradient characteristics; Based on the thermal damage degree and energy accumulation state of each regulation zone, the partition energy compensation coefficient is generated; the high-frequency current output is differentiated according to the partition energy compensation coefficient until the thermodynamic response signal meets the global inactivation condition.

Images