A method and system for controlling the travel speed of physical weeding equipment

Abstract

This invention relates to the field of weeding equipment control, and particularly to a method and system for controlling the travel speed of physical weeding equipment. It includes: acquiring signals of the spatial distribution of the work area to be worked on and constructing a parameterized task space model; extracting the projection vector relationship of weed feature clusters within the motion boundary constraints of the work space; determining the frequency of spatial task execution based on the projection vector relationship, and determining the minimum task cycle by combining real-time data throughput rate and physical response rate; determining a time delay term based on the installation distance between the forward sensing module and the controlled operating system and the real-time feed rate; and generating a feed rate control command based on the minimum task cycle and the time delay term to adjust the travel speed. Through feedforward spatial topology pre-sensing and kinematic constraint calculation, adaptive speed adjustment for non-uniform weed distribution is achieved, avoiding actuator saturation and logic accumulation, and improving weeding accuracy and work efficiency.

Description

Technical Field

[0001] This invention relates to the field of weeding equipment control, and in particular to a method, system, electronic device, and computer-readable storage medium for controlling the travel speed of physical weeding equipment. Background Technology

[0002] Currently, in the automated operation process of precision agriculture, target identification and targeted removal are achieved in complex farmland environments by integrating visual sensing units and physical execution terminals. In existing automated weeding operations, a constant-rate operation mode is usually set. Due to the non-uniform and random distribution of weeds in the field, this fixed control logic causes the system to face mutually restrictive operational contradictions in different load areas. When the equipment enters a high-density weed area, the number of targets to be processed surges per unit time. The sensing sampling frequency of the main control system and the response speed of the end effector will touch the physical limit of the control system, resulting in instruction accumulation, logical conflicts and execution deviations in the control loop. From the perspective of system control theory, this feedback control based on real-time density has obvious phase lag. If a low-speed travel strategy is adopted to ensure accuracy, a lot of invalid operation time will be generated in sparse weed areas, which will significantly reduce the overall operational efficiency of the system.

[0003] Besides physical architecture limitations, lagging control logic is a core bottleneck restricting operational efficiency. For example, existing technologies disclose a laser weeding method and laser weeding machine that adaptively adjusts laser dosage. This scheme adjusts laser power and operating speed according to the growth stage of weed species. The adjustment logic is a discrete matching based on specific target classification, which fails to reach the spatial topological distribution relationship of the operating target. When the equipment enters a non-uniform clustered distribution area, due to the lack of calculation of load distribution topological geometric constraints, it is difficult to quantify the physical response boundary of the end effector in handling high-density lattice based solely on single target classification and identification. This induces action conflicts or logic accumulation of the end effector, resulting in lag in perception feedback and physical execution response. The speed adjustment process lacks mathematical kinematic continuity considerations, and frequent large-scale speed switching weakens the stability of operation, generates instantaneous current surges, and exacerbates mechanical fatigue of the drive system. Summary of the Invention

[0004] To address the problems of lagging travel speed regulation, actuator saturation, and decreased operational accuracy caused by the lack of spatial topology pre-sensing and kinematic constraints in existing physical weeding equipment under non-uniform weed distribution conditions, this invention provides a method, system, electronic device, and computer-readable storage medium for controlling the travel speed of physical weeding equipment.

[0005] In a first aspect, the present invention provides a method for controlling the travel speed of a physical weeding device, comprising:

[0006] The spatial distribution signal of the task area is collected, spatial features are extracted from the collected spatial distribution signal, and a parameterized task space model is constructed using the extracted spatial features; the parameterized task space model is used to characterize the topological distribution signal of the task load.

[0007] Extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space;

[0008] Based on the projection vector relationship, the spatial task execution frequency of the end effector mapping module when covering weed feature clusters is determined, and the minimum task cycle for the controlled operation system to process the area to be worked is determined in combination with the operating parameters; the operating parameters include the real-time data throughput rate of the controlled operation system and the physical response rate of the end effector mapping module, the controlled operation system includes at least the end effector mapping module, and the end effector mapping module stores preset work space motion boundary constraints;

[0009] The time delay term for the controlled operation system to reach the work area is determined based on the installation distance between the forward-looking perception data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment; the forward-looking perception data preprocessing module is used to collect the work space distribution signal.

[0010] A feed rate control command is generated based on the minimum task cycle and the time delay term, and the travel rate of the physical weeding device is adjusted based on the feed rate control command.

[0011] Optionally, extracting the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space includes:

[0012] Extract the topological coordinates of each weed feature point contained in the weed feature cluster in the three-dimensional spatial coordinate system from the parameterized task space model;

[0013] The topological coordinates are projected onto the two-dimensional motion vector plane of the end effector mapping module in the feed direction to establish the geometric interference lattice of the weed feature clusters relative to the motion boundary constraints of the work space, so as to obtain the projection vector relationship.

[0014] Optionally, determining the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship includes:

[0015] Calculate the distribution density and spatial gradient of the geometric interference lattice on the sweep path of the motion boundary constraint of the work space;

[0016] Based on the distribution density and the spatial gradient, the action switching state of the end effector mapping module within a unit feed distance is determined, and the rate of change of the action switching state is used as the execution frequency of the spatial task.

[0017] Optionally, determining the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship, and determining the minimum task cycle for the controlled operation system to process the area to be worked in combination with operating parameters, includes:

[0018] The number of weed feature points contained in the weed feature clusters within the work area and the physical displacement travel corresponding to each weed feature point are determined based on the frequency of execution of the spatial task.

[0019] The maximum motion rate constrained by the workspace motion boundary and the inherent response delay of the end effector mapping module when logically switching between adjacent feature points are determined based on the physical response rate of the end effector mapping module.

[0020] The formula for calculating the minimum task cycle is:

[0021] ;

[0022] Among them, T min The minimum task cycle is given by n, where n is the number of feature points in the weed feature clusters within the area to be worked on, and L is the minimum task cycle. i V is the physical displacement travel required for the end effector mapping module to process the i-th weed feature point. max Δτ represents the maximum motion rate constrained by the workspace motion boundary, and Δτ represents the inherent response delay of the end effector mapping module when logically switching between adjacent feature points. The data transmission and computation delay is determined based on the real-time data throughput rate of the controlled operating system and the data volume of the parameterized task space model.

[0023] Optionally, generating the feed rate control command based on the minimum task cycle and the time delay term includes:

[0024] Obtain the feed rate gradient of the preceding and subsequent operation cycles adjacent to the area to be worked;

[0025] The feed rate gradient is used as the first derivative constraint of the spline function at the corresponding time node, and the spline function is used to fit the discrete rate control point corresponding to the minimum task cycle to generate a feed rate control command that maintains the smoothness of the acceleration vector on the time axis.

[0026] Optionally, the controlled operation system further includes a pre-detection signal acquisition unit, and the method further includes:

[0027] Receive the missed operation sampling data fed back by the pre-detection signal acquisition unit;

[0028] The task weight coefficient in the minimum task cycle calculation process is dynamically adjusted based on the missed operation sampling data; the task weight coefficient is used to correct the minimum task cycle so as to maintain the operation accuracy within a preset deviation range during the feed rate change process.

[0029] Optionally, the controlled operation system further includes a feed drive unit monitoring unit, and the method further includes:

[0030] Monitor the instantaneous feedback current of the feed drive unit;

[0031] When the speed change caused by the feed rate control command results in the instantaneous feedback current change gradient exceeding a preset load protection threshold, the physical weeding device is kept within a preset operating power range by limiting the absolute value of the rate gradient corresponding to the feed rate control command.

[0032] In a second aspect, the present invention provides a physical weeding equipment travel speed control system, wherein the physical weeding equipment travel speed control system is applied to the physical weeding equipment travel speed control method described in any one of the first aspects, and the physical weeding equipment travel speed control system comprises:

[0033] The forward-looking perception data preprocessing module is used to collect work space distribution signals in the work area, extract spatial features from the collected work space distribution signals, and construct a parameterized task space model using the extracted spatial features. The parameterized task space model is used to characterize the task load topology distribution signal.

[0034] The central rate adjustment module is used to extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space;

[0035] A central rate adjustment module is used to determine the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship, and to determine the minimum task cycle for the controlled operation system to process the area to be worked in combination with the operating parameters; the controlled operation system includes at least the end effector mapping module, and the end effector mapping module stores preset work space motion boundary constraints;

[0036] The central rate adjustment module is used to determine the time delay term for the controlled operation system to reach the work area based on the installation distance between the forward sensing data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment; the forward sensing data preprocessing module is used to collect the work space distribution signal.

[0037] The central rate adjustment module is used to generate feed rate control commands based on the minimum task cycle and the time delay term;

[0038] A controlled operating system for adjusting the travel rate of the physical weeding equipment based on the feed rate control command.

[0039] Thirdly, the present invention provides an electronic device, comprising: a memory for storing instructions; and a processor for invoking the instructions stored in the memory to execute the physical weeding device travel rate control method as described in any one of the first aspects.

[0040] Fourthly, the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, perform the physical weeding device travel rate control method as described in any one of the first aspects.

[0041] To address the problems of lag in travel speed adjustment, actuator saturation, and decreased operational accuracy caused by the lack of spatial topology pre-sensing and kinematic constraints in existing physical weeding equipment under non-uniform weed distribution conditions, this invention has the following advantages:

[0042] In physical weeding equipment, the environmental perception state of the unoperated area is acquired in real time through the front vision module, and the central control unit converts the weed distribution into a parameterized point cloud model composed of multiple centroid coordinates. Combined with the kinematic constraint model of the end operation module, the feed profile constraint is solved, so that the system can complete the rate pre-compensation before the weeds enter the main operation area. This realizes the leap from passive response to active adaptive optimization. By introducing a spatial delay term compensation algorithm, the response lag and regulation oscillation of the control system when dealing with non-uniform loads are eliminated, and the feed continuity in complex topology environments is guaranteed from the perspective of system stability.

[0043] By using parameterized operation templates to structure the topology of weed distribution, the minimum time consumption of the actuator to avoid trajectory logic conflicts when covering a specific point cloud model is calculated. From the geometric constraint level, the system predicts and solves the problem of missed identification and missed operation caused by motion congestion of the end effector. This enables the system to maintain a high feed rate while ensuring sufficient computation and execution redundancy in high-load areas, thus achieving a synergistic improvement in operation efficiency and operation accuracy.

[0044] The dynamic adjustment mechanism based on pre-test design generates a feed rate curve with mathematical continuity, avoiding frequent start-stop and large-span acceleration / deceleration switching of the drive motor when facing instantaneous changes in weed density. This reduces instantaneous current surges and mechanical transmission losses in the drive system, improves the operating status of the vehicle drive module, and extends the mechanical lifespan of the power system. By constructing a projection correlation between the virtual motion envelope and the parameterized point cloud model, the system establishes a digital solution coupling interface between the control platform and the end-user module, constructing a typical cloud-edge-end collaborative or distributed control logic architecture. This makes the control logic no longer limited to specific physical weeding methods, forming a universal task space feed planning and control algorithm. By updating the kinematic constraint model of the actuator, it can achieve rapid adaptation to different weeding terminals and generation of optimal feed strategies, enhancing the versatility and engineering application flexibility of the operating platform in heterogeneous actuators and diverse agronomic scenarios. Attached Figure Description

[0045] Figure 1 A schematic flowchart of a method for controlling the travel speed of a physical weeding device according to some embodiments is shown;

[0046] Figure 2 This diagram illustrates the intelligent travel speed control principle and data flow of physical weeding equipment according to some embodiments;

[0047] Figure 3 The diagram illustrates the overall logical architecture and information interaction topology of some embodiments;

[0048] Figure 4 A diagram of a travel rate control device for a physical weeding apparatus according to some embodiments is shown;

[0049] Figure 5 A schematic diagram of an electronic device is shown.

[0050] Figure label:

[0051] Physical weeding equipment travel speed control system 20; forward sensing data preprocessing module 21; central speed regulation module 22; controlled operation system 23; electronic equipment 400; memory 401; processor 402; input / output (I / O) interface 403. Detailed Implementation

[0052] The invention will now be discussed with reference to several exemplary embodiments. It should be understood that these embodiments are described merely to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0053] As used herein, the term "comprising" and its variations are to be interpreted as open-ended terms meaning "including but not limited to". The term "based on" is to be interpreted as "at least partially based on". The terms "one embodiment" and "an embodiment" are to be interpreted as "at least one embodiment". The term "another embodiment" is to be interpreted as "at least one other embodiment".

[0054] Example 1:

[0055] This embodiment discloses a method for controlling the travel speed of a physical weeding device proposed in this invention, used for controlling the travel speed of the physical weeding device, such as... Figure 1 As shown, the process may include steps S10 to S50, which will be explained in detail below:

[0056] In this embodiment, the execution entity of the physical weeding equipment travel speed control method is the control system of the physical weeding equipment. The control system includes a forward-looking sensing data preprocessing module, a central speed regulator, and a controlled operation system. The controlled operation system includes a pre-detection signal acquisition unit, a feed drive unit, and an end effector mapping module.

[0057] This embodiment provides an intelligent travel speed control system for a physical weeding device, comprising a forward-looking perception data preprocessing module, a controlled operating system, and a central speed regulator. Each module achieves distributed data interaction through a standardized digital logic interface, constructing a control topology with strong real-time performance and high reliability. Essentially, this system is an adaptive closed-loop control architecture oriented towards complex spatiotemporal constraints. By transforming environmental excitation signals into a parameterized task space model, it achieves independent solution and dynamic correction of the control law. The forward-looking perception data preprocessing module is electrically connected to the central speed regulator and is used to collect and construct the task space distribution signal of the area to be worked. A parameterized task space model is used. The controlled operating system includes a pre-detection signal acquisition unit, a feed drive unit, and an end effector mapping module for performing image sampling and physical weeding actions. A central rate regulator is connected to both the look-ahead perception data preprocessing module and the controlled operating system. Through the execution of regulation logic, it calculates the projection vector relationship of weed feature clusters within the movement boundary constraints of the operating space. Combining the system's real-time data throughput rate and physical response rate, it determines the minimum task cycle and generates a feed rate control command that satisfies the first-order derivative continuity constraint, outputting it to the feed drive unit. This achieves the coupling of perception, decision-making, and execution, as well as adaptive dynamic adjustment of the travel rate. The intelligent travel rate control principle and data flow of the physical weeding equipment can be referenced... Figure 2 As shown, the overall logical architecture and information interaction topology involved in the data processing process can be referenced. Figure 3 As shown.

[0058] This embodiment combines Figures 2 to 3 Description of the intelligent travel speed control system for physical weeding equipment, such as... Figure 2 As shown, the environmental input of the work area provides the system with a work space distribution signal. The forward-looking perception data preprocessing module collects the work space distribution signal of the work area, calculates the geometric constraint parameters of the work space and constructs a parameterized task space model, and transmits it to the central rate regulator. The central rate regulator receives the real-time data throughput rate reference from the pre-detection signal acquisition unit, and at the same time, combined with the work space motion boundary constraints provided by the end effector mapping module, calculates the projection vector relationship of weed feature clusters and determines the minimum task cycle and spatial delay term, thereby generating a feed rate control command that satisfies the first derivative continuity constraint, and sends the command to the feed drive unit in the controlled work system. The controlled work system also includes a pre-detection signal acquisition unit for assisting real-time data acquisition, and an end effector mapping module that stores preset work space motion boundary constraints and provides physical response rate parameters.

[0059] like Figure 3 As shown, the overall architecture of the system is presented as follows: the physical host of the physical weeding equipment constructs the physical carrier boundary, which contains a forward-looking perception data preprocessing module as a visual computing unit. It receives the input of the area to be operated and outputs a parameterized task space model to the central rate regulator, which is the core decision computing node. The central rate regulator interacts with the upper control platform, which is the remote management terminal, through bidirectional logical bit streams, and sends feed rate control commands to the controlled operating system, which is the driving and execution unit. During operation, the controlled operating system feeds back closed-loop correction requests to the central rate regulator, thus forming a complete control closed loop.

[0060] Step S10: Collect the spatial distribution signal of the work area, extract the spatial features of the collected spatial distribution signal, and use the extracted spatial features to construct a parameterized task space model.

[0061] In this embodiment, the forward-looking perception data preprocessing module acquires 3D point cloud data or depth images of the area to be worked on using spatial location sensing components (such as binocular cameras or LiDAR), obtaining the original spatial distribution signal of the work area containing weeds, crops, and the ground. Then, geometric feature operators (such as segmentation algorithms based on shape, color, or height) are used to identify the centroid coordinates of individual weeds from the original signal. Each weed centroid is abstracted as a feature point, which typically includes its lateral, longitudinal, and depth coordinates in a 3D spatial coordinate system, as well as an optional attribute identifier. All extracted weed feature points are aggregated according to their spatial relationships to form a parametric point cloud model, i.e., a parametric task space model. This model directly represents the topological distribution of the task load (weeds) within the work area, including weed density, cluster morphology, relative distances between feature points, and gradient changes.

[0062] To facilitate direct access by the central rate regulator, the coordinates and attributes of each feature point are encoded using a fixed bit width (e.g., 128 bits), with necessary header redundancy factored in, and finally encapsulated into a quantizable logical bitstream. The central rate regulator does not need to understand the original image content; it only needs to read the point cloud model to determine the distribution density and spatial gradient of the weeds. The specific calibration logic for calculating the data volume is as follows: multiply the number of identified weed feature points by the fixed storage bit width of 128 bits, which includes 32 bits of horizontal coordinates, 32 bits of vertical coordinates, 32 bits of depth coordinates, and 32 bits of attribute identifiers, and factor in a header redundancy factor of 1.2. Divide the resulting total number of bits by 1,000,000 to convert it into a value in Mb units.

[0063] Step S20: Extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the task space.

[0064] In this embodiment, the central rate regulator retrieves the parameterized task space model and extracts the topological coordinates of each weed feature point contained in the weed feature cluster in the three-dimensional spatial coordinate system. Then, each topological coordinate is projected onto the two-dimensional motion vector plane of the end effector mapping module in the feed direction, thereby establishing a geometric interference lattice of the weed feature cluster relative to the workspace motion boundary constraints. This geometric interference lattice is the desired projection vector relationship. The workspace motion boundary constraints are pre-stored in the end effector mapping module to limit the motion range of the actuator.

[0065] Preferably, extracting the projection vector relationship of the weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space includes: extracting the topological coordinates of each weed feature point contained in the weed feature cluster in the three-dimensional spatial coordinate system from the parameterized task space model; projecting the topological coordinates onto the two-dimensional motion vector plane of the end effector mapping module in the feed direction, and establishing the geometric interference lattice of the weed feature cluster relative to the motion boundary constraints of the work space to obtain the projection vector relationship.

[0066] It should be noted that the central rate regulator executes the feed rate generation logic. The regulator retrieves the parameterized task space model, extracts the projection vector relationship of weed feature clusters in the model within the work space motion boundary constraints, and extracts the topological coordinates of each weed feature point in the three-dimensional spatial coordinate system. It then projects these coordinates onto the two-dimensional motion vector plane of the end effector mapping module in the feed direction to establish a geometric interference lattice of weed feature clusters relative to the work space motion boundary constraints. By calculating the distribution density and spatial gradient of the geometric interference lattice on the sweep path of the work space motion boundary constraints, the regulator determines the action switching state of the end effector mapping module within a unit feed distance. The specific mapping logic is as follows: the central rate regulator divides the spatial grid along the travel direction in 5mm increments, counts the number of grids containing weed feature points within a continuous 100mm displacement distance, multiplies this count by the 0.05s inherent response time required for the end effector to complete a single reciprocating weeding action, and obtains the total action execution time for this interval. The ratio of this time to the total displacement time is defined as the rate of change of the action switching state.

[0067] Step S30: Based on the projection vector relationship, determine the spatial task execution frequency of the end effector mapping module when covering weed feature clusters, and combine the operating parameters to determine the minimum task cycle for the controlled operation system to process the area to be operated.

[0068] In this embodiment, the operating parameters include the real-time data throughput rate of the controlled operating system and the physical response rate of the end effector mapping module. The controlled operating system includes at least the end effector mapping module, which stores preset workspace motion boundary constraints.

[0069] Preferably, determining the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship includes: calculating the distribution density and spatial gradient of the geometric interference lattice on the sweep path constrained by the motion boundary of the work space; determining the action switching state of the end effector mapping module within a unit feed distance based on the distribution density and the spatial gradient, and using the rate of change of the action switching state as the spatial task execution frequency.

[0070] Specifically, the spatial grid is divided along the direction of travel in 5mm increments. The number of grids containing weed feature points within a continuous 100mm displacement distance is counted. This count is multiplied by the inherent response time required for the end effector to complete a single reciprocating weeding action to obtain the total execution time of the action in that interval. The ratio of this to the total displacement time is defined as the rate of change of the action switching state. This rate of change is used as the frequency of spatial task execution.

[0071] Preferably, the spatial task execution frequency of the end effector mapping module when covering weed feature clusters is determined based on the projection vector relationship, and the minimum task cycle for the controlled operation system to process the area to be worked is determined in combination with the operating parameters. This includes: determining the number of feature points of weed feature points contained in the weed feature clusters in the area to be worked and the physical displacement travel corresponding to each weed feature point according to the spatial task execution frequency; and determining the maximum movement rate limited by the movement boundary constraints of the work space and the inherent response delay of the end effector mapping module when logically switching between adjacent feature points according to the physical response rate of the end effector mapping module.

[0072] The formula for calculating the minimum task cycle is:

[0073] ;

[0074] Among them, T min The minimum task cycle is given by n, where n is the number of feature points in the weed feature clusters within the area to be worked on, and L is the minimum task cycle. i V is the physical displacement travel required for the end effector mapping module to process the i-th weed feature point. max Δτ represents the maximum motion rate constrained by the workspace motion boundary, and Δτ represents the inherent response delay of the end effector mapping module when logically switching between adjacent feature points. The data transmission and computation delay is determined based on the real-time data throughput rate of the controlled operating system and the data volume of the parameterized task space model.

[0075] It should be understood that, to prevent the end effector from experiencing execution saturation when handling high-load areas, the central rate regulator calculates the minimum task cycle for the controlled job system to process the area to be worked. This calculation process is based on the projection vector relationship and combines the real-time data throughput rate of the controlled job system and the physical response rate of the end effector mapping module. The central rate regulator obtains the real-time data throughput rate B of the data bus in real time. rate And determine the amount of solution data S of the parameterized task space model within the current field of view. data By calculating the amount of data S data With real-time data throughput rate B rate The ratio determines the computational delay component T of the system.comp =S data / B rate And will calculate the delay component T comp Superimposed on the computational model of the minimum task cycle, where B rate The real-time effective bandwidth of the system communication link, measured in Mbps. data T represents the amount of geometric and topological information contained in the weed feature clusters in a single solution task, expressed in Mb. In the formula for calculating the minimum task cycle above, T... min The minimum task cycle is given by n, where n is the number of feature points in the weed feature clusters within the area to be worked on, and L is the minimum task cycle. i The physical displacement travel required for the end effector mapping module to process the i-th weed feature point, in units of V max The maximum motion speed is constrained by the workspace motion boundary, expressed in mm / s. Δτ is the inherent response delay of the end effector mapping module when logically switching between adjacent feature points. The data transmission and computation delay is determined based on the real-time data throughput rate of the controlled operating system and the data volume of the parameterized task space model, and is expressed in seconds.

[0076] Step S40: Determine the time delay term for the controlled operation system to reach the work area based on the installation distance between the forward sensing data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment.

[0077] In this embodiment, when determining physical property parameters, the system uses hardware procedures to determine the maximum motion speed V. max The quantization setting with the inherent response delay Δτ includes the following steps: placing the end effector mapping module in calibration mode, controlling the actuator to perform reciprocating stepping motion across the entire range, and collecting real-time displacement curves using an absolute encoder integrated into the motor shaft. The maximum slope of the displacement curve is recorded as V. max , where V max The unit is mm / s; the timing of the control pulse transmission and the triggering time of the limit switch of the mechanical component are monitored using a high-speed data acquisition card. The average of 100 time differences between the two is defined as Δτ, where Δτ is stored in non-volatile memory and participates in the minimum task cycle T. min Discrete solution eliminates response deviations caused by the mechanical inertia of the hardware terminal; the central rate regulator determines the response based on the installation distance D between the look-ahead sensing data preprocessing module and the controlled operating system, as well as the real-time feed rate V of the physical weeding equipment. current Determine the time delay term T for the controlled operating system to reach the work area. delayTo address the sensor optical axis deflection caused by operations on slopes, the system collects the pitch angle feedback from the inertial measurement unit in real time. The nominal 1200mm installation spacing is multiplied by the cosine correction factor of this angle to calculate the real-time effective horizontal projection distance. This value is then used as the correction input parameter for calculating the time delay term T. delay Real-time updates rely on installation spacing D measurement and feed rate feedback. The method for determining the feed rate includes: using a laser alignment instrument to measure the linear projection distance (D) from the optical center of the vision sensor in the look-ahead sensing data preprocessing module to the execution geometric center of the end effector mapping module on the horizontal reference plane; and using the laser alignment instrument to measure the linear projection distance (D) from the optical center of the vision sensor in the look-ahead sensing data preprocessing module to the execution geometric center of the end effector mapping module on the horizontal reference plane. The unit of D is mm. The central rate regulator reads the data from the Hall sensor built into the feed drive unit to calculate the current feed rate V. current According to formula T delay =D / V current The real-time spatial delay is calculated, and the rate gradient of adjacent work cycles is used as a boundary constraint. A feed rate control command is generated by solving a system of cubic spline equations. This ensures that the current change rate of the drive motor is limited within the hysteresis protection threshold of the power converter when crossing boundary regions with different weed densities, achieving a smooth transition of travel speed across spatial and temporal scales. Based on the time delay term and the minimum task cycle, the central rate regulator generates the feed rate control command. To achieve smooth adjustment of the feed rate, the regulator obtains the feed rate gradients of the preceding and subsequent work cycles and uses spline functions to fit the discrete rate control points corresponding to the minimum task cycle. The specific execution procedure is as follows: a cubic spline interpolation algorithm with an discrete step length of 10ms is used. The current feed rate and the gradient of the preceding feed rate are used as the first derivative constraint of the initial segment. Fitting coefficients are generated by solving a system of three equations for rate and acceleration at consecutive time points, ensuring that the acceleration change at the connection point of each command segment is less than 0.10 m / s². 2 This ensures that the generated feed rate control command satisfies the first derivative continuity constraint, maintains the continuity of the acceleration vector on the time axis, and avoids instantaneous large current surges in the drive motor.

[0078] Step S50: Generate a feed rate control command based on the minimum task cycle and the time delay term, and adjust the travel rate of the physical weeding device based on the feed rate control command.

[0079] In this embodiment, after determining the minimum task cycle and the time delay term, the rate gradient of adjacent work cycles is retrieved as a boundary constraint condition. The feed rate control command is generated by solving a system of cubic spline equations. This ensures that the rate of change of the current of the drive motor is limited within the hysteresis protection threshold of the power converter when crossing boundary areas with different weed densities, thus achieving a smooth transition of the travel speed in the spatiotemporal scale. Based on the time delay term and the minimum task cycle, the central rate regulator generates the feed rate control command. To achieve smooth adjustment of the feed rate, the regulator obtains the feed rate gradient of the preceding and subsequent work cycles and uses spline functions to fit the discrete rate control points corresponding to the minimum task cycle.

[0080] Preferably, generating a feed rate control command based on the minimum task cycle and the time delay term includes: obtaining the feed rate gradients of the preceding and subsequent task cycles adjacent to the work area; using the feed rate gradients as first-order derivative constraints of the spline function at the corresponding time nodes, and fitting the discrete rate control points corresponding to the minimum task cycle using the spline function to generate a feed rate control command that maintains smooth acceleration vector on the time axis.

[0081] It should be noted that the central rate regulator acquires the feed rate gradients of the preceding and following work cycles adjacent to the work area, and uses these feed rate gradients as the first derivative constraints of the spline function at the corresponding time nodes. A cubic spline interpolation algorithm with an offset step size of 10ms is used, retrieving the current feed rate and the gradient of the preceding feed rate as the first derivative constraints of the initial segment, with the minimum task period T... min The corresponding target velocity point serves as the end constraint. By solving the three-variable equations of velocity and acceleration at consecutive time points, fitting coefficients are generated to ensure that the acceleration change at the connection of each command segment is less than a preset threshold (e.g., 0.10 m / s²). This generates a feed rate control command that maintains a smooth acceleration vector on the time axis. The command is then output to the feed drive unit to adjust the travel rate of the physical weeding device.

[0082] Preferably, the end effector mapping module includes a kinematic equivalent model of at least one of a laser working head, a mechanical component, and a high-frequency pulse generator. Specifically, the end effector mapping module provides adaptability to heterogeneous working terminals, including a kinematic equivalent model of at least one of a laser working head, a mechanical component, or a high-frequency pulse generator. By updating the internally stored kinematic feature operators, the system adapts to the motion constraints of different actuators in the feed direction.

[0083] Preferably, the controlled operating system is also used to execute the following closed-loop feedback logic: real-time monitoring of the logic execution redundancy of the end effector mapping module, and when the logic execution redundancy is lower than a preset 5% threshold, sending a deceleration adjustment request to the central rate regulator to perform closed-loop correction of the feed rate control command.

[0084] Preferably, the central rate regulator is also used to perform the following decoupling logic: by establishing a digital logic interface between the controlled operating system and the upper control platform, the parameterized task space model is mapped into a logic bit stream that can be read by the end-effector mapping module, so as to achieve decoupling between the control logic and the physical actuator.

[0085] It should be noted that by establishing a digital logical interface between the controlled operating system and the upper-level control platform, the parameterized task space model is mapped into a logical bitstream readable by the end-effector mapping module, achieving coordination between control logic and physical actuators. The controlled operating system executes closed-loop feedback logic to cope with dynamic interference. The system monitors the logical execution redundancy of the end-effector mapping module in real time. When it falls below a preset 5% threshold, the controlled operating system sends a deceleration adjustment request to the central rate regulator to correct the feed rate control command. The central rate regulator monitors the logical execution redundancy of the end-effector mapping module and corrects the rate command accordingly. The logical execution redundancy is defined as R. logic The calculation method includes: within the complete task cycle of the controlled operation system, using a timer to collect the idle time T after the laser head or mechanical component completes a single weeding action. free and the total duration T of the task cycle total , will R logic Set as T free With T total The ratio; R was detected logic When the value is below the preset limit of 5%, the central speed regulator reduces the duty cycle of the control signal output to the feed drive unit, thereby lowering the travel speed of the physical weeding equipment to R. logic The temperature has risen back to a stable range, avoiding the risk of actuator overload and thermal overload caused by dense weed distribution.

[0086] Preferably, the central rate regulator simultaneously monitors the instantaneous feedback current of the feed drive unit. When the current change gradient exceeds a preset load protection threshold, the regulator limits the absolute value of the rate gradient to stabilize the system within a preset operating power range. Specifically, this includes monitoring the instantaneous feedback current of the feed drive unit. When the speed change caused by the feed rate control command results in the instantaneous feedback current change gradient exceeding the preset load protection threshold, the regulator limits the absolute value of the rate gradient corresponding to the feed rate control command to keep the physical weeding device within a preset operating power range. The preset operating power range can be set according to actual needs while ensuring safety, and is not limited thereto.

[0087] Preferably, the regulator receives missed operation sampling data fed back by the pre-detection signal acquisition unit and dynamically adjusts the task weight coefficient in the calculation process, including receiving the missed operation sampling data fed back by the pre-detection signal acquisition unit; dynamically adjusting the task weight coefficient in the minimum task cycle calculation process according to the missed operation sampling data, wherein the task weight coefficient is used to correct the minimum task cycle so as to maintain the operation accuracy within a preset deviation range during the feed rate change process.

[0088] It should be noted that the initial calibration value of the weight coefficient is 1.00. The system counts the total number of missed operation feature points fed back by the pre-detection signal acquisition unit in the last 10 operation cycles in real time. If the missed operation rate is higher than 2%, the weight coefficient is adjusted upward with a step gradient of 0.10 to the maximum upper limit of 1.50, thereby increasing the feedback gain in the calculation logic of the minimum task cycle. Conversely, it is adjusted downward with a gradient of 0.05 to maintain the operation accuracy within the preset deviation range.

[0089] Further explanation will be provided in the following application scenarios:

[0090] In specific agricultural operations, when physical weeding equipment enters an area where weeds are distributed in non-uniform clusters and the local topological density momentarily exceeds the processing capacity of the controlled operation system, the forward-looking perception data preprocessing module acquires the centroid coordinates of the weeds in real time through the spatial location sensing component and constructs a parameterized task space model. For this specific topological structure, if the system relies solely on real-time sampling feedback for adjustment, the spatial delay term T between perception and execution will be insufficient. delay This causes logical stacking and mechanical impact when the end effector mapping module reaches the weed feature cluster.

[0091] The central rate regulator retrieves the aforementioned model, extracts the projection vector relationship of weed feature clusters within the motion boundary constraints of the workspace, maps the weed target points into a geometric interference lattice on a two-dimensional motion vector plane, and then applies the formula... The minimum task cycle is determined. The feed drive unit receives the feed rate control command generated by the central rate regulator, which satisfies the first derivative continuity constraint. It then uses a spline function to smoothly fit the discrete rate control points, ensuring the feed rate V... current Before reaching that area, smoothly descend to the minimum mission cycle T. min Matching values, this feedforward adjustment method based on geometric constraints transforms the speed control logic into a dynamic constraint solution based on spatial topology, enabling the system to maintain within the preset operational accuracy deviation range when dealing with complexly distributed weed clusters, and to keep the mechanical load of the drive system stable.

[0092] To address the issue of missed operations caused by sensor response lag and actuator interference in the operation of physical weeding equipment under non-uniform weed clusters, this experiment constructed a semi-physical simulation platform integrating a look-ahead sensing preprocessing module and a controlled operation system. The platform verifies the adjustment characteristics of the adaptive travel speed control system. A spatial distribution signal set containing random pitting noise is used as the input source to simulate the objective environment where the spatial position sensing component is affected by dust and drastic changes in light. The measurement resolution of the spatial position sensing component is set to 1 mm, the sampling frequency to 20 Hz, the speed regulation resolution of the feed drive unit to be no less than 0.01 m / s, the core parameter installation spacing D to 1200 mm, the minimum physical response delay Δτ to 0.05 s, and the maximum movement speed V... max The installation spacing D is set to 800 mm / s. The value of the installation spacing D is used to balance the time margin of look-ahead preprocessing with the compactness of the overall structure. When D decreases, the buffer time of the system to deal with abrupt feature clusters is shortened, and the corresponding calculation frequency of the central rate regulator is increased.

[0093] The experiment was conducted in three phases: data injection, logic calculation, and result mapping. In the data injection phase, Gaussian random displacement disturbances with a mean of 0 and a standard deviation of 5 mm were superimposed onto the parameterized task space model to simulate the mechanical vibration deviation of the end effector when operating on uneven ground. In the logic calculation phase, the central rate regulator retrieved noisy topological coordinates in real time and performed projection vector relationship calculations to determine the geometric interference lattice distribution of weed feature clusters on the two-dimensional motion vector plane. The minimum task period T was then determined according to the formula. min In the result mapping stage, the feed drive unit executes the feed rate control command generated by spline function fitting, and records the operation accuracy and system response curves under different weed distribution densities. To verify the synergistic effect and the rationality of the parameter range, a control group with a fixed rate, a comparison group without a prospective perception preprocessing module, and an experimental group using the scheme of this invention are set up.

[0094] Table 1: Example table of system performance comparison data under different weed topological complexities

[0095]

[0096] Analysis of the data in Table 1 shows that in a high-density clustered distribution scenario with 35 weed feature points (n), the fixed-rate control group experienced a missed operation rate of 28.7% because the actuator could not complete the entire physical displacement stroke within a unit time. The non-forward-looking perception control group reduced the rate through real-time feedback adjustment, but due to the spatial delay T between perception and execution... delay This resulted in the speed control command taking effect after missing the target area, with a missed operation rate of 8.5% and a drive current fluctuation rate increasing to 18.4%. The test group using the present invention calculated the projection vector relationship and pre-generated control commands that satisfied the first derivative continuity constraint, thus smoothly reducing the feed rate V before entering the high-density area. current This keeps the missed operation rate below 1.5% and the driving current fluctuation rate stable at 2.8%. This data confirms the role of spatial topology modeling and feedforward control mechanism in resolving the contradiction between perception delay and execution saturation. When the number of weed feature points n increases to 60 and the distribution is extremely dense, the system maintains the operation accuracy by reducing the feed rate to 0.05m / s, but the operation area per unit time decreases. The physical response rate of the controlled operation system approaches the kinematic limit. At this time, the logic execution redundancy monitored by the controlled operation system in real time is lower than the 5% threshold, triggering the closed-loop correction logic of the central rate regulator. By limiting the rate gradient of the feed rate control command, the system can operate stably at extremely low speed.

[0097] In scenarios involving uneven, undulating terrain, the spatial positioning sensing component experiences optical axis deflection due to the ground slope. The look-ahead sensing data preprocessing module corrects the visual data using a built-in attitude compensation operator. This process extracts the homogeneous transformation matrix M fed back from the external inertial measurement unit. pose and the collected original coordinates P raw After multiplication and post-processing, it is transformed to a reference operating coordinate system based on the feed direction of the physical weeding equipment, where M pose To characterize the sensor attitude deviation, P is a 4×4 homogeneous transformation matrix. raw Based on the collected three-dimensional spatial discrete coordinate vectors, the central rate regulator performs discrete calculation of the projection vector relationship, extracts the set of three-dimensional coordinate points of the weed centroid in the reference working coordinate system, and uses a projection operator perpendicular to the feed vector to reduce its dimension and map it to the dynamic working plane of the end effector. The regulator uses a gridded spatial sampling window to perform lattice segmentation on the projected two-dimensional plane, where the grid size is set to 5mm×5mm. When the number of projection points in the sampling window is greater than 1, the system marks the window as a geometric interference feature point, and then constructs a geometric interference lattice that reflects the mutual exclusivity of the action logic of the actuator.

[0098] The load protection threshold of the feed drive unit is determined by the following physical characteristic calibration procedure: The reference operating current I of the feed drive unit is collected under no-load conditions. base The feed rate V is increased in increments of 0.05 m / s. current Upon reaching the rated value, the average steady-state current at each rate point is recorded. The central rate regulator monitors the proportional relationship between the feedback current change gradient ΔI and the feed rate change gradient ΔV in real time. When the proportional relationship deviates from the preset impedance linear range by 15%, the system determines the current change at this critical node as the load protection threshold. The quantification procedure for this threshold is as follows: Under the no-load operation state of the equipment, a linear proportional reference relationship of 0.15A feedback current corresponding to every 0.01m / s rate increment is established by gradually increasing the feed rate. If the real-time monitored current change rate exceeds the 15% offset of this reference ratio, it is determined that the current gradient exceeds the load protection threshold. Here, ΔI is the current increment of the feed drive unit in A, and ΔV is the change in feed rate in m / s. The feed rate control command is generated according to the third-order spline interpolation procedure. The central rate regulator retrieves the current feed rate V. current With the acceleration vector A at the current moment current As the starting endpoint constraint of the interpolation interval, and with the minimum task period T min The corresponding target rate point is used as the endpoint constraint. The coefficient matrix of the rate fitting curve is determined by constructing and solving a system of continuity equations, where A current The real-time acceleration of the feed drive unit, in m / s². 2 This procedure controls the electromagnetic torque pulsation of the drive motor below the demagnetization threshold of the magnet, so that the controlled operating system can maintain the dynamic stability of the operating feed when faced with the impact of discrete feature clusters.

[0099] The end effector mapping module retrieves the kinematic equivalent model of the laser working head or the kinematic equivalent model of the mechanical components based on the physical attribute identifier code of the currently mounted terminal. The central rate regulator extracts the feature parameter f representing the terminal's execution frequency from the selected model. act and the characteristic parameter R that characterizes the effective working radius eff And substitute it into the minimum task cycle T min In the computational logic, f act R is the inherent operating frequency of the actuator, measured in Hz. eff The effective coverage radius of the end effector is in mm. When the current end effector is identified as a mechanical weeding claw with a low response frequency, the feature operator increases the value of the minimum physical response delay Δτ, so that the generated feed rate control command is within the time delay term T. delayCorresponding downward adjustments are generated at the nodes. This adaptation process enables the controlled operating system to achieve a balance between operational efficiency and accuracy within different physical constraint boundaries. By determining the minimum task cycle and performing rate fitting, the system ensures that operational accuracy remains within a preset deviation range even when the topological complexity of weed distribution fluctuates drastically.

[0100] When replacing hardware components in a physical weeding device, the installation height H and installation spacing D of the spatial position sensing component undergo physical changes. The system executes a pre-calibration procedure to determine the projection transformation reference for the parameterized task space model. This procedure includes placing a calibration object on a horizontal reference ground, measuring the horizontal displacement from the optical center of the spatial position sensing component to the motion center of the end effector and assigning it as the installation spacing D, the central rate regulator acquiring the three-axis gravity components of the inertial measurement unit in a static state and calculating the initial attitude angle, and constructing a homogeneous transformation matrix M for coordinate correction. pose This allows the raw coordinates P collected by the forward-looking perception data preprocessing module to be displayed. raw When transformed to the reference working coordinate system, it matches the physical execution space.

[0101] Under operating conditions where the dynamic characteristics of the feed drive units differ, the central speed regulator executes the calibration process for the inherent response delay Δτ. The system controls the feed drive unit to move at the rated feed rate and triggers the action switching command. The time difference between the control signal issuance time and the sampling feedback time is recorded. The average time of 100 consecutive action switchings is calculated and determined as the minimum physical response delay Δτ. Then, combined with the installation spacing D and the real-time feed rate V, the system further optimizes the response delay. current Determine the time delay term T delay The real-time solution benchmark, where T delay The time delay term, measured in seconds, synchronizes the sensor's physical position with the electronic control response hysteresis characteristics of the end effector, ensuring that the generated feed rate control command is executed within the time delay term T. delay After buffering, it maintains a spatiotemporal correspondence with the weed topology of the area to be worked on.

[0102] When physical weeding equipment faces high-density and complex vegetation distribution, the spatial resolution G of the system's task execution is limited due to the uncertainty in the physical size and distribution density of the feature clusters to be processed. resThe calibration procedure involves placing a calibration matrix within the look-ahead field of view, consisting of feature spheres with a diameter of 10 mm and a geometric center spacing of 30 mm. The central rate regulator controls the controlled operating system to pass through the calibration area at a rated speed. The look-ahead sensing data preprocessing module acquires the spatial distribution signal of the task area and extracts the projected geometric interference lattice. The central rate regulator calculates the centroid coincidence of adjacent feature points. When the centroid coincidence is less than 1% and the spatial position deviation is within the measurement resolution of 1 mm, the system determines the current grid size as the task spatial resolution G. res Among them, G res The side length of the gridded spatial sampling window is in mm. This procedure quantifies and correlates the physical characteristics of the weed topology with the modeling accuracy of the parameterized task space model, so that the generated geometric task description has a definite spatial resolution.

[0103] When the central rate regulator and the remote control platform exchange data, fluctuations in the communication link load will cause a time delay term T. delay The dynamic offset, the system execution communication delay Δt com The real-time monitoring and compensation process involves the central rate regulator periodically sending probe data packets with hardware time stamps to the controlled operating system. It measures the time difference between the issuance of the command and the return of an execution confirmation signal from the controlled operating system. The system calculates the arithmetic mean of 500 consecutive observations within a preset sliding time window and determines this as the communication delay Δt. com , where Δt com The logical latency of the communication link, measured in seconds, is Δt, which is the communication delay generated by the central rate regulator. com With the physical time delay term T delay Perform superposition calculations and adjust the feed rate control command based on the feedback of missed operation sampling data to make the feed rate V... current When faced with communication interference, it remains within a stable range constrained by the movement boundaries of the workspace.

[0104] This embodiment acquires the spatial distribution signal of the work area to be worked and constructs a parameterized task space model; extracts the projection vector relationship of weed feature clusters within the motion boundary constraints of the work space; determines the spatial task execution frequency based on the projection vector relationship, and determines the minimum task cycle by combining the real-time data throughput rate and the physical response rate of the end effector; determines the time delay term based on the installation distance between the forward sensing module and the controlled work system and the real-time feed rate; and generates a feed rate control command that satisfies the first-order derivative continuity constraint based on the minimum task cycle and the time delay term, thereby adjusting the travel rate. This invention achieves adaptive rate adjustment for non-uniform weed distribution through feedforward spatial topology pre-sensing and kinematic constraint solving, avoiding actuator saturation and logic accumulation, and improving weeding accuracy and work efficiency.

[0105] Those skilled in the art will understand that the above embodiments are specific examples of implementing this disclosure, and in practical applications, various changes can be made in form and detail without departing from the scope of this disclosure.

[0106] Example 2:

[0107] Based on the same inventive concept, this disclosure also provides a physical weeding equipment travel speed control system 20, such as... Figure 4 As shown, the travel speed control system 20 of the physical weeding equipment includes:

[0108] The forward-looking perception data preprocessing module 21 is used to collect the work space distribution signal of the work area, extract the spatial features of the collected work space distribution signal, and construct a parameterized task space model using the extracted spatial features. The parameterized task space model is used to characterize the task load topology distribution signal.

[0109] Central speed adjustment module 22 is used to extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space;

[0110] The central rate adjustment module 22 is used to determine the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship, and to determine the minimum task cycle for the controlled operation system to process the area to be worked in combination with the operating parameters; the controlled operation system includes at least the end effector mapping module, and the end effector mapping module stores preset work space motion boundary constraints;

[0111] The central rate adjustment module 22 is used to determine the time delay term for the controlled operation system to reach the work area based on the installation distance between the forward sensing data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment; the forward sensing data preprocessing module is used to collect the work space distribution signal.

[0112] The central rate adjustment module 22 is used to generate a feed rate control command based on the minimum task cycle and the time delay term.

[0113] The controlled operating system 23 is used to adjust the travel rate of the physical weeding equipment based on the feed rate control command.

[0114] This embodiment acquires the spatial distribution signal of the work area to be worked and constructs a parameterized task space model; extracts the projection vector relationship of weed feature clusters within the motion boundary constraints of the work space; determines the spatial task execution frequency based on the projection vector relationship, and determines the minimum task cycle by combining the real-time data throughput rate and the physical response rate of the end effector; determines the time delay term based on the installation distance between the forward sensing module and the controlled work system and the real-time feed rate; and generates a feed rate control command that satisfies the first-order derivative continuity constraint based on the minimum task cycle and the time delay term, thereby adjusting the travel rate. This invention achieves adaptive rate adjustment for non-uniform weed distribution through feedforward spatial topology pre-sensing and kinematic constraint solving, avoiding actuator saturation and logic accumulation, and improving weeding accuracy and work efficiency.

[0115] Example 3:

[0116] Based on the same inventive concept, this disclosure also provides a computer-readable storage medium storing instructions that, when executed by a processor, perform the physical weeding device travel speed control method as described in any one of the first aspects.

[0117] Example 4:

[0118] like Figure 5 As shown, one embodiment of this disclosure provides an electronic device 400. The electronic device 400 includes a memory 401, a processor 402, and an input / output (I / O) interface 403. The memory 401 stores instructions. The processor 402 executes the physical weeding device travel rate control method of this disclosure by calling the instructions stored in the memory 401. The processor 402 is connected to both the memory 401 and the input / output (I / O) interface 403, for example, via a bus system and / or other forms of connection mechanisms (not shown). The memory 401 can be used to store programs and data, including the program for the physical weeding device travel rate control method involved in the embodiments of this disclosure. The processor 402 executes various functional applications and data processing of the electronic device 400 by running the program stored in the memory 401.

[0119] In this embodiment of the disclosure, the processor 402 may be implemented in at least one of the following hardware forms: digital signal processor (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 402 may be one or a combination of several of the following: central processing unit (CPU) or other processing units with data processing capability and / or instruction execution capability.

[0120] The memory 401 in this embodiment may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD).

[0121] In this embodiment of the disclosure, the input / output (I / O) interface 403 can be used to receive input instructions (such as numeric or character information, and key signal inputs related to user settings and function control of the electronic device 400), and can also output various information (such as images or sounds) to the outside. In this embodiment of the disclosure, the input / output (I / O) interface 403 may include one or more of the following: a physical keyboard, function keys (such as volume control keys, power buttons, etc.), a mouse, a joystick, a trackball, a microphone, a speaker, and a touch panel.

[0122] It is understood that although operations are described in a specific order in the accompanying drawings in the embodiments of this disclosure, this should not be construed as requiring these operations to be performed in the specific order or serial order shown, or requiring all of the shown operations to be performed to obtain the desired result. In certain environments, multitasking and parallel processing may be advantageous.

[0123] The methods and apparatus disclosed herein can be implemented using standard programming techniques, and various method steps can be implemented using rule-based logic or other logic. It should also be noted that the terms "apparatus" and "module" as used herein and in the claims are intended to include implementations using one or more lines of software code and / or hardware implementations and / or devices for receiving input.

[0124] Any step, operation, or procedure described herein may be performed or implemented using one or more hardware or software modules, either alone or in combination with other devices. In one embodiment, the software module is implemented using a computer program product comprising a computer-readable medium containing computer program code, which is executable by a computer processor to perform any or all of the described steps, operations, or procedures.

[0125] The foregoing description of embodiments of this disclosure has been provided for purposes of illustration and description. The foregoing description is not exhaustive and is not intended to limit this disclosure to the exact form disclosed; various modifications and variations may be made in accordance with the foregoing teachings, or may be derived from practice of this disclosure. These embodiments were chosen and described to illustrate the principles of this disclosure and its practical application, enabling those skilled in the art to utilize this disclosure in various implementations and modifications suitable for the particular purpose conceived.

Claims

1. A method for controlling the travel speed of a physical weeding device, characterized in that, The method for controlling the travel speed of the physical weeding equipment includes: The spatial distribution signal of the task area is collected, spatial features are extracted from the collected spatial distribution signal, and a parameterized task space model is constructed using the extracted spatial features; the parameterized task space model is used to characterize the topological distribution signal of the task load. Extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space; Based on the projection vector relationship, the spatial task execution frequency of the end effector mapping module when covering weed feature clusters is determined, and the minimum task cycle for the controlled operation system to process the area to be worked is determined in combination with the operating parameters; the operating parameters include the real-time data throughput rate of the controlled operation system and the physical response rate of the end effector mapping module, the controlled operation system includes at least the end effector mapping module, and the end effector mapping module stores preset work space motion boundary constraints; The time delay term for the controlled operation system to reach the work area is determined based on the installation distance between the forward-looking perception data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment; the forward-looking perception data preprocessing module is used to collect the work space distribution signal. A feed rate control command is generated based on the minimum task cycle and the time delay term, and the travel rate of the physical weeding device is adjusted based on the feed rate control command.

2. The method for controlling the travel speed of physical weeding equipment according to claim 1, characterized in that, The step of extracting the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space includes: Extract the topological coordinates of each weed feature point contained in the weed feature cluster in the three-dimensional spatial coordinate system from the parameterized task space model; The topological coordinates are projected onto the two-dimensional motion vector plane of the end effector mapping module in the feed direction to establish the geometric interference lattice of the weed feature clusters relative to the motion boundary constraints of the work space, so as to obtain the projection vector relationship.

3. The method for controlling the travel speed of physical weeding equipment according to claim 2, characterized in that, The determination of the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship includes: Calculate the distribution density and spatial gradient of the geometric interference lattice on the sweep path of the motion boundary constraint of the work space; Based on the distribution density and the spatial gradient, the action switching state of the end effector mapping module within a unit feed distance is determined, and the rate of change of the action switching state is used as the execution frequency of the spatial task.

4. The method for controlling the travel speed of physical weeding equipment according to claim 1, characterized in that, The step of determining the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship, and determining the minimum task cycle for the controlled operation system to process the area to be worked in combination with the operating parameters, includes: The number of weed feature points contained in the weed feature clusters within the work area and the physical displacement travel corresponding to each weed feature point are determined based on the frequency of execution of the spatial task. The maximum motion rate constrained by the workspace motion boundary and the inherent response delay of the end effector mapping module when logically switching between adjacent feature points are determined based on the physical response rate of the end effector mapping module. The formula for calculating the minimum task cycle is: ； Among them, T min The minimum task cycle is given by n, where n is the number of feature points in the weed feature clusters within the area to be worked on, and L is the minimum task cycle. i V is the physical displacement travel required for the end effector mapping module to process the i-th weed feature point. max Δτ represents the maximum motion rate constrained by the workspace motion boundary, and Δτ represents the inherent response delay of the end effector mapping module when logically switching between adjacent feature points. The data transmission and computation delay is determined based on the real-time data throughput rate of the controlled operating system and the data volume of the parameterized task space model.

5. The method for controlling the travel speed of a physical weeding device according to claim 1, characterized in that, The generation of feed rate control instructions based on the minimum task cycle and the time delay term includes: Obtain the feed rate gradient of the preceding and subsequent operation cycles adjacent to the area to be worked; The feed rate gradient is used as the first derivative constraint of the spline function at the corresponding time node, and the spline function is used to fit the discrete rate control point corresponding to the minimum task cycle to generate a feed rate control command that maintains the smoothness of the acceleration vector on the time axis.

6. The method for controlling the travel speed of a physical weeding device according to claim 1, characterized in that, The controlled operation system further includes a pre-detection signal acquisition unit, and the method further includes: Receive the missed operation sampling data fed back by the pre-detection signal acquisition unit; The task weight coefficient in the minimum task cycle calculation process is dynamically adjusted based on the missed operation sampling data; the task weight coefficient is used to correct the minimum task cycle so as to maintain the operation accuracy within a preset deviation range during the feed rate change process.

7. The method for controlling the travel speed of a physical weeding device according to any one of claims 1 to 6, characterized in that, The controlled operating system further includes a feed drive monitoring unit, and the method further includes: Monitor the instantaneous feedback current of the feed drive unit; When the speed change caused by the feed rate control command results in the instantaneous feedback current change gradient exceeding a preset load protection threshold, the physical weeding device is kept within a preset operating power range by limiting the absolute value of the rate gradient corresponding to the feed rate control command.

8. A travel speed control system for a physical weeding device, characterized in that, The physical weeding equipment travel speed control system is applied to the physical weeding equipment travel speed control method as described in any one of claims 1 to 7, wherein the physical weeding equipment travel speed control system comprises: The forward-looking perception data preprocessing module is used to collect work space distribution signals in the work area, extract spatial features from the collected work space distribution signals, and construct a parameterized task space model using the extracted spatial features. The parameterized task space model is used to characterize the task load topology distribution signal. The central rate adjustment module is used to extract the projection vector relationship of weed feature clusters in the parameterized task space model within the motion boundary constraints of the work space; The central rate adjustment module is used to determine the spatial task execution frequency of the end effector mapping module when covering weed feature clusters based on the projection vector relationship, and to determine the minimum task cycle for the controlled operation system to process the area to be operated in combination with the operating parameters; the controlled operation system includes at least the end effector mapping module, and the end effector mapping module stores preset work space motion boundary constraints; The central rate adjustment module is used to determine the time delay term for the controlled operation system to reach the work area based on the installation distance between the forward sensing data preprocessing module and the controlled operation system and the real-time feed rate of the physical weeding equipment; the forward sensing data preprocessing module is used to collect the work space distribution signal. The central rate adjustment module is used to generate feed rate control commands based on the minimum task cycle and the time delay term; A controlled operating system for adjusting the travel rate of the physical weeding equipment based on the feed rate control command.

9. An electronic device, characterized in that, include: Memory, used to store instructions; And a processor, used to invoke instructions stored in the memory to execute the physical weeding device travel rate control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The device stores instructions that, when executed by a processor, perform the physical weeding device travel speed control method as described in any one of claims 1 to 7.

Images