An intelligent control system for spraying pressure of a spraying rod of a plant protection robot
By monitoring and adjusting the spray pressure in real time on the spray rod of the plant protection robot, the spraying problem of uneven spraying caused by the spray head is solved, and the uniformity and efficiency of spraying is improved, and it is adapted to complex operating conditions and meteorological environment.
Patent Information
- Application Number
- CN202411209879.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-08-30
AI Technical Summary
The spray heads of existing plant protection robot spray rods are prone to blockage when spraying for a long time and the pesticide concentration is high, resulting in uneven spraying range. Traditional systems cannot monitor and adjust spray pressure in real time, affecting crop protection effect and pesticide utilization efficiency.
A spray pressure intelligent control system for spray rods of plant protection robots is designed, including an initial pressure setting module and a real-time pressure adjustment module. The spray image is captured using a high-speed camera, and the spray diffusion range is simulated through morphological analysis and dynamic model, real-time monitoring and Adjust the spray pressure to ensure that the coverage between the nozzles is completely overlapping.
The uniformity and efficiency of spraying are improved, the utilization rate of pesticides is maximized, the negative environmental impact is reduced, the reliability and durability of the system are enhanced, and the adaptability to various operating conditions and meteorological environments are adopted.
Smart Images

Figure CN118732724B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of pressure control systems, and in particular to an intelligent control system for spraying pressure of a spraying rod of a plant protection robot. Background Art
[0002] In the existing field of plant protection robots, the nozzles on the spraying booms often face the problem of clogging when the spraying operation is carried out for a long time and the pesticide concentration is high, which will cause blind spots in the spraying range, thus affecting the uniformity and effectiveness of field spraying. Usually, the spraying range between the nozzles needs to be pre-calibrated to achieve full coverage, but partial clogging of the nozzles will destroy this uniform coverage, resulting in insufficient spraying in some areas of the field, thereby affecting the effect of crop protection and the efficiency of pesticide use.
[0003] In traditional systems, the spray pressure is usually pre-set and will not be adjusted according to the actual situation during the spraying process. This fixed-pressure spraying method lacks flexibility and cannot deal with problems often encountered in actual operations such as nozzle blockage. In addition, traditional systems rarely monitor the spraying range and nozzle status in real time, making it difficult to immediately discover and solve the problem of uneven spraying.
[0004] Therefore, it is necessary to develop a new type of intelligent control system for spraying pressure of the spraying boom of the plant protection robot, which can monitor the spraying status and the working condition of the nozzle in real time and intelligently adjust the spraying pressure to ensure comprehensive and uniform coverage of the field even when the nozzle is partially blocked. Summary of the invention
[0005] Based on the above purpose, the present invention provides an intelligent control system for spraying pressure of a spraying rod of a plant protection robot.
[0006] An intelligent control system for spraying pressure of a spraying rod of a plant protection robot, comprising an initial pressure setting module and a real-time pressure adjustment module;
[0007] The initial pressure setting module sets the initial spraying pressure according to the distance between adjacent spray heads of the spraying rod, so that the spraying range between every two adjacent spray heads is fully covered;
[0008] During the actual spraying operation, the real-time pressure adjustment module monitors the spraying coverage between two adjacent nozzles in real time to determine whether the coverage is complete. If the coverage is incomplete, the spraying pressure is adjusted; if the coverage is complete, no adjustment is made; the real-time pressure adjustment module also includes:
[0009] Spray pattern capture submodule: uses a high-speed camera to capture the umbrella-shaped spray image formed by the pesticide sprayed from the nozzle;
[0010] Morphological analysis submodule: Identify the boundary and structure of umbrella-shaped spray through edge detection and morphological analysis;
[0011] Diffusion range prediction submodule: uses a kinetic model to simulate the diffusion path and terminal distribution of spray particles to obtain the spray range;
[0012] Pressure adjustment logic submodule: Based on the spraying range, calculate the spraying pressure compensation value and adjust the spraying pressure to correct the deviation.
[0013] Furthermore, the initial pressure setting module adjusts the spray pressure in the spray rod by operating the total pressure valve, observes the spraying range between each two adjacent nozzles, and continues to adjust the total pressure valve until it is confirmed that the spraying range of each adjacent nozzle is completely overlapped.
[0014] Furthermore, the spray pattern capturing submodule uses at least one high-speed camera installed on the spray rod, aimed at the spray outlet of at least one nozzle, to capture the umbrella-shaped spray image formed by the sprayed pesticide in real time, and record the image of the pesticide from spraying to initial diffusion.
[0015] Furthermore, the morphological analysis submodule specifically includes:
[0016] Image preprocessing: The spray images captured by the high-speed camera were first preprocessed, including denoising, contrast enhancement, and color correction;
[0017] Edge detection: Use edge detection algorithm to process the preprocessed image, identify the boundary line of the spray, reveal the geometric shape and external contour of the spray, and provide a basis for understanding the spray range;
[0018] Morphological analysis: Based on edge detection, morphological analysis is performed, including erosion and dilation processing, to eliminate noise and fill in the discontinuities in the spray image;
[0019] Spray feature extraction: Based on the results of edge detection and morphological analysis, key spray features are extracted, including the maximum diameter of the spray, the average diffusion angle, the particle size distribution and the diffusion velocity.
[0020] Furthermore, the edge detection algorithm adopts the Canny edge detection algorithm, which is as follows:
[0021] Gaussian filter: To adapt to the umbrella-shaped spray image, adjust the standard deviation of the Gaussian filter To capture the width and edge fluidity of the spray, it is calculated as:
[0022] ,in, is the Gaussian filter for position Response, The value is selected based on the average size of the spray diameter to ensure that the spray edge remains clear after filtering;
[0023] Gradient calculation and direction: Calculate the gradient strength and direction of the image to determine the boundary. The gradient strength is calculated as follows:
[0024] ;in, and They are the gradients of the image in the x and y directions, and the gradient directions for: ;
[0025] Non-maximum suppression: For spray images, when performing non-maximum suppression, the algorithm is adjusted to adapt to the curved shape of the spray to ensure that the edges are detected along the natural curve of the spray;
[0026] Hysteresis threshold: Use two thresholds to binarize the gradient image and determine the true edge points. If the gradient value is higher than the high threshold, it is considered an edge point; if the gradient value is lower than the low threshold, it is excluded; if the gradient value is between the two thresholds, it is retained only if it is connected to the edge point.
[0027] Further, the morphological analysis: morphological operations are used to analyze and process image structures, including corrosion and expansion:
[0028] The erosion operation is used to remove details from the image boundary and is defined by the following formula:
[0029] ,in, is the original image, is a structural element, Is a structural element Pan to Position the situation at that time;
[0030] The dilation operation is used to enhance features in an image and is defined as:
[0031] ,in, represents the intersection, which means that as long as and exist The locations overlap. belongs to the result image.
[0032] Furthermore, in the non-maximum suppression, for each pixel, its neighboring pixels in the gradient direction are checked. If the intensity of the neighboring pixel in the gradient direction is not the largest, its intensity is set to 0, and only the pixel with the local maximum intensity is retained to achieve edge refinement. The specific steps are as follows:
[0033] For each pixel in the image, according to its gradient direction , find two adjacent pixels in this direction;
[0034] Compare the gradient strength of the current pixel With the gradient strength of two adjacent pixels, if the gradient strength of the current pixel is not the largest, the pixel is considered not to belong to the edge and its strength is set to 0;
[0035] Repeat the above process until all pixels of the entire image have been processed.
[0036] Furthermore, the diffusion range prediction submodule specifically includes:
[0037] Receiving morphological analysis data: The diffusion range prediction submodule receives the spray image data from the morphological analysis submodule, including spray boundary structure, angle, diameter and particle distribution information;
[0038] Dynamic model application: Based on the received data, a customized fluid dynamics model is used to simulate the movement and diffusion behavior of the spray particles in the air. Initial conditions are set in the customized fluid dynamics model. The boundary structure of the spray is used to set the initial velocity field and pressure field of the customized fluid dynamics model. The customized fluid dynamics model includes the Navier-Stokes equations, the spray particle behavior model, and the environmental condition interaction model.
[0039] The NavierStokes equations are expressed as:
[0040] ,in, is the velocity field, is the density, It's pressure. is the kinematic viscosity, Represents the resistance caused by the spray particle size and distribution, is the external force term, i.e. the influence of environmental wind on the spray;
[0041] The spray particle behavior model considers the movement and interaction of a single spray particle or a particle group, and uses the following equation to represent the motion state of the particle:
[0042] ;
[0043] ;
[0044] in, and are the position and velocity of the particle, is the relaxation time of the particle, which indicates the time required for the particle velocity to adjust to the airflow velocity, is the local air velocity, is the acceleration due to gravity;
[0045] The environmental condition interaction model is used to describe the behavior of the spray under the influence of external conditions by changing the external force term in the NavierStokes equation. To simulate the effects of wind:
[0046] ,in, is the air density, is the drag coefficient related to particle shape and size, is the wind speed vector, is the velocity vector of the particle or fluid;
[0047] The diffusion range prediction submodule also includes simulating the diffusion path and terminal distribution and spraying range prediction, wherein the simulated diffusion path and terminal distribution are simulated by numerically solving the NavierStokes equation, simulating the movement path of the spray particles starting from the nozzle until reaching the terminal distribution, and calculating the position and speed of the spray particles at different time and space points, thereby predicting the overall diffusion range of the spray; the spraying range prediction determines the actual coverage area of the spray based on the simulation results.
[0048] Furthermore, the numerical solution of the Navier-Stokes equation is as follows:
[0049] Meshing: Divide the computational domain into grids or elements so that the equations can be solved in each small area;
[0050] Discretization equations: The continuous Navier-Stokes equations are discretized using the finite volume method and converted into a system of equations that can be solved on a computer. For each control volume, the Navier-Stokes equations are integrated to obtain:
[0051] ,in, represents the control volume, , , , represent velocity, pressure, density and kinematic viscosity respectively, represents the spray particle resistance, Represents a tiny volume element. It is used to represent an infinitesimal volume portion within the control volume during integration. When integrating fluid dynamics equations, it is necessary to integrate the fluid properties within the control volume (CV). Used to represent volume elements in these integrals;
[0052] Solving discrete equations: Apply iterative solvers to solve the discretized equations to obtain approximate solutions of flow field parameters such as velocity and pressure. Through the iterative process, the true values of flow field parameters such as velocity and pressure are gradually approached;
[0053] Particle tracking simulation: Use the obtained flow field data to simulate the movement path of the spray particles, and use the flow field data to perform particle tracking simulation, specifically: initialize the particle position and velocity; in each time step, according to the flow field velocity and external forces (such as gravity, wind force, etc.) at the current particle position, update the particle velocity and position:
[0054] ;
[0055] ;in, and are the velocity and position of the particle at time t, is the total force acting on the spray particles, is the particle mass, is the time step.
[0056] The spraying range prediction specifically includes:
[0057] Path integration: Establish the overall diffusion model of the spray by integrating all the particle paths obtained by simulation;
[0058] Terminal distribution analysis: Evaluate the terminal distribution of particles in time and space to determine the coverage area and shape of the spray.
[0059] Furthermore, the pressure adjustment logic submodule specifically includes:
[0060] Determine the insufficient coverage range of spraying: Assume that the complete coverage range of adjacent sprinklers is , the actual coverage is ,if , then there is insufficient coverage;
[0061] Calculate the undercoverage ratio: Undercoverage ratio Calculated by the following formula:
[0062] ;
[0063] Relating spray pressure to spray range: Spray pressure And spraying range The linear relationship between them is expressed as: ,in, is the coefficient of relationship between spray range and pressure, is a constant;
[0064] Calculate pressure compensation: To achieve full coverage, determine the amount of pressure increase that will compensate for the lack of coverage , it is deduced that: ; Solve this equation to get the required pressure compensation value: ; or based on the undercoverage ratio To calculate: ;
[0065] Adjust the spray pressure: adjust the current spray pressure Add the calculated pressure compensation value , get the new spray pressure: ; Apply new spray pressure , adjust the spraying range to achieve ideal full coverage effect.
[0066] Beneficial effects of the present invention:
[0067] The present invention can accurately control the spraying range of each nozzle by real-time monitoring of the spraying range and automatically adjusting the spraying pressure, ensuring that the spraying areas between adjacent nozzles completely overlap. This fine control mechanism significantly improves the uniformity of spraying, reduces missed spraying and repeated spraying, maximizes the utilization rate of pesticides, and reduces the potential negative impact on the environment. If the pressure of multiple nozzles on the spray rod is individually controlled, it can ensure that even if a nozzle is partially blocked, the pressure of adjacent nozzles can be adjusted to compensate for the coverage range and maintain the uniformity of spraying. Such a dynamic adjustment mechanism not only improves the uniformity of spraying, but also enhances the reliability and durability of the system, and reduces the risk of reduced spraying efficiency and insufficient crop protection due to nozzle blockage.
[0068] The present invention, by integrating a morphology analysis submodule and a diffusion range prediction submodule, can respond to changes in spray morphology and fluctuations in external environmental conditions in real time. This dynamic adaptation mechanism enables the system to maintain an optimized spraying effect under various operating conditions, thereby improving operating efficiency and adaptability, and maintaining efficient and stable spraying performance even under complex field conditions and unstable meteorological environments.
[0069] The present invention provides data-driven decision support for spray pressure adjustment through the spray image captured by a high-speed camera and the numerical solution of the fluid dynamics model. This intelligent control system can analyze the spray state and environmental changes in real time, automatically adjust the pressure, and ensure precise control of the spray range. BRIEF DESCRIPTION OF THE DRAWINGS
[0070] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0071] Figure 1 This is a schematic diagram of the functional modules of the intelligent control system according to an embodiment of the present invention;
[0072] Figure 2 Schematic diagram of a diffusion range prediction submodule according to an embodiment of the present invention. DETAILED DESCRIPTION
[0073] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.
[0074] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0075] like Figure 1-2 As shown, an intelligent control system for spraying pressure of a spraying rod of a plant protection robot includes an initial pressure setting module and a real-time pressure adjustment module;
[0076] The initial pressure setting module sets the initial spraying pressure according to the distance between adjacent nozzles of the spraying rod, so that the spraying range between every two adjacent nozzles is fully covered;
[0077] During the actual spraying operation, the real-time pressure adjustment module monitors the spraying coverage between two adjacent nozzles in real time to determine whether the coverage is complete. If the coverage is incomplete, the spraying pressure is adjusted; if the coverage is complete, no adjustment is made. The real-time pressure adjustment module also includes:
[0078] Spray pattern capture submodule: uses a high-speed camera to capture the umbrella-shaped spray image formed by the pesticide sprayed from the nozzle;
[0079] Morphological analysis submodule: Identify the boundary and structure of umbrella-shaped spray through edge detection and morphological analysis;
[0080] Diffusion range prediction submodule: uses a kinetic model to simulate the diffusion path and terminal distribution of spray particles to obtain the spray range;
[0081] Pressure adjustment logic submodule: Based on the spraying range, calculate the spraying pressure compensation value and adjust the spraying pressure to correct the deviation.
[0082] The initial pressure setting module adjusts the spray pressure in the spray rod by operating the total pressure valve, observes the spray range between each two adjacent nozzles, and continues to adjust the total pressure valve until it is confirmed that the spray range of each adjacent nozzle is completely overlapped;
[0083] If the pressures of multiple nozzles on the spraying boom are individually controlled, the initial pressure setting module also includes adjusting the pressure setting of each nozzle by individually operating it.
[0084] The spray pattern capture submodule uses at least one high-speed camera installed on the spray rod, aimed at the spray outlet of at least one nozzle, to capture the umbrella-shaped spray image formed by the sprayed pesticide in real time, and record the image of the pesticide from spraying to initial diffusion.
[0085] The morphological analysis submodule specifically includes:
[0086] Image preprocessing: The spray images captured by the high-speed camera were first preprocessed, including denoising, contrast enhancement, and color correction;
[0087] Edge detection: Use edge detection algorithm to process the preprocessed image, identify the boundary line of the spray, reveal the geometric shape and external contour of the spray, and provide a basis for understanding the spray range;
[0088] Morphological analysis: Based on edge detection, morphological analysis is performed, including erosion and dilation processing, to eliminate noise and fill the gaps in the spray image. Through morphological operations, the overall structure and internal characteristics of the spray can be better understood, such as the uniformity of the spray and the distribution of particle density;
[0089] Spray feature extraction: Based on the results of edge detection and morphological analysis, key spray features are extracted, including the maximum diameter of the spray, the average diffusion angle, the particle size distribution and the diffusion velocity.
[0090] The edge detection algorithm uses the Canny edge detection algorithm. For umbrella-shaped spray images, the edge is usually formed on the periphery of the spray, so we need to pay special attention to the outer edge of the spray and its shape and continuity, as follows:
[0091] Gaussian filter: To adapt to the umbrella-shaped spray image, adjust the standard deviation of the Gaussian filter To capture the width and edge fluidity of the spray, it is calculated as:
[0092] ,in, is the Gaussian filter for position Response, The value is selected based on the average size of the spray diameter to ensure that the spray edge remains clear after filtering;
[0093] Gradient calculation and direction: Calculate the gradient strength and direction of the image to determine the boundary. The gradient strength is calculated as follows:
[0094] ;in, and They are the gradients of the image in the x and y directions, and the gradient directions for: ;
[0095] Non-maximum suppression: For spray images, when performing non-maximum suppression, the algorithm is adjusted to adapt to the curved shape of the spray to ensure that the edges are detected along the natural curve of the spray;
[0096] Hysteresis threshold: Use two thresholds to binarize the gradient image and determine the true edge points. If the gradient value is higher than the high threshold, it is considered an edge point; if the gradient value is lower than the low threshold, it is excluded; if the gradient value is between the two thresholds, it is retained only if it is connected to the edge point.
[0097] Morphological analysis: Morphological operations are used to analyze and manipulate image structures, including erosion and dilation:
[0098] The erosion operation is used to remove details from the image boundary and is defined by the following formula:
[0099] ,in, is the original image, is a structural element, Is a structural element Pan to Position the situation at that time;
[0100] The dilation operation is used to enhance features in an image and is defined as:
[0101] ,in, represents the intersection, which means that as long as and exist The locations overlap. belongs to the result image;
[0102] In the present invention, edge detection is used to accurately identify the boundaries of the umbrella-shaped spray, thereby determining the coverage and shape of the spray, and morphological analysis further processes these boundaries, removes noise points and fills gaps to obtain more accurate and complete spray morphology information.
[0103] In non-maximum suppression, for each pixel, its neighboring pixels in the gradient direction are checked. If the intensity of the neighboring pixel in the gradient direction is not the largest, its intensity is set to 0, and only the pixels with the local maximum intensity are retained to achieve edge refinement. The specific steps are as follows:
[0104] For each pixel in the image, according to its gradient direction , find two adjacent pixels in that direction. For example, if the gradient direction is horizontal (0 degrees or 180 degrees), check the left and right pixels; if the gradient direction is vertical (90 degrees or 270 degrees), check the top and bottom pixels.
[0105] Compare the gradient strength of the current pixel With the gradient strength of two adjacent pixels, if the gradient strength of the current pixel is not the largest, the pixel is considered not to belong to the edge and its strength is set to 0;
[0106] Repeat the above until all pixels of the entire image have been processed;
[0107] The specific calculation is as follows:
[0108] set up For point The gradient strength at is the gradient direction of the point;
[0109] Non-maximum suppression can be expressed by the following logic:
[0110] For each pixel ,according to Determine the direction of comparison.
[0111] set up and For The gradient strength of two neighboring pixels in the direction.
[0112] if Greater than and ,but remain unchanged, otherwise Set to .
[0113] Mathematically it can be expressed as:
[0114] = ;in is the gradient strength after non-maximum suppression.
[0115] In this way, we ensure that only pixels with local maximum intensity in the gradient direction of each pixel are considered edges. In this way, the result after non-maximum suppression is a refined edge image, in which the edge represents the center line of the actual physical boundary, which is crucial for subsequent edge tracking and analysis. When processing umbrella-shaped spray images, this refined edge is more helpful to accurately capture the boundary and shape of the spray.
[0116] The diffusion range prediction submodule specifically includes:
[0117] Receiving morphological analysis data: The diffusion range prediction submodule receives the spray image data from the morphological analysis submodule, including spray boundary structure, angle, diameter and particle distribution information;
[0118] Application of dynamic model: Based on the received data, a customized fluid dynamic model is used to simulate the movement and diffusion behavior of spray particles in the air. Initial conditions are set in the customized fluid dynamic model. The boundary structure based on the spray is used to set the initial velocity field and pressure field of the customized fluid dynamic model. For example, if the morphological analysis shows that the spray forms a specific umbrella-shaped distribution, then this distribution can be simulated in the initial conditions of the model to ensure that the simulation process starts from a physically accurate initial state of the spray. The boundary structure also helps define the conditions on the boundary of the fluid model. For example, the identification results of the spray boundary can be used to define the velocity distribution and pressure conditions on the flow field boundary to ensure that the model boundary reflects the physical behavior of the actual spray. This means that on the outer boundary of the spray, the customized fluid dynamic model will be adjusted to match the actual extension and diffusion velocity of the spray. The customized fluid dynamic model includes the Navier-Stokes equations, the spray particle behavior model, and the environmental condition interaction model;
[0119] The Navier-Stokes equations are expressed as:
[0120] ,in, is the velocity field, is the density, It's pressure. is the kinematic viscosity, Represents the resistance caused by the spray particle size and distribution, is the external force term, i.e. the influence of environmental wind on the spray;
[0121] The spray particle behavior model considers the movement and interaction of a single spray particle or a group of particles, and uses the following equation to represent the motion state of the particles:
[0122] ;
[0123] ;
[0124] in, and are the position and velocity of the particle, is the relaxation time of the particle, which indicates the time required for the particle velocity to adjust to the airflow velocity, is the local air velocity, is the acceleration due to gravity;
[0125] While solving the Navier-Stokes equations to obtain the velocity field and pressure distribution of the fluid, the spray particle behavior model is applied to this flow field to track and simulate the motion of individual spray particles. When calculating the fluid dynamics, the model also calculates the behavior of the particles in the flow field, including phenomena such as velocity changes, collisions, and aggregation. The behavior of the particles (such as changes in velocity and direction) will be fed back to the fluid dynamics model, especially in areas where the particle concentration is high or the particle interaction significantly affects the flow field. This feedback is achieved by adjusting the velocity and pressure distribution in the flow field, thereby reflecting the impact of particle behavior on fluid dynamics.
[0126] The environmental condition interaction model is used to describe the behavior of the spray under the influence of external conditions by changing the external force terms in the Navier-Stokes equations. To simulate the effects of wind:
[0127] ,in, is the air density, is the drag coefficient related to particle shape and size, is the wind speed vector, is the velocity vector of particles or fluids. This model can be adjusted according to actual meteorological data to reflect the impact of environmental conditions such as wind on spray behavior in real time;
[0128] The environmental condition interaction model is mainly realized by adjusting the boundary conditions and external force terms in the Navier-Stokes equation. For example, the wind impact can be simulated by modifying the external force terms in the equation, and the boundary condition adjustment of temperature and humidity changes can simulate the influence of thermodynamic effects on spray behavior. The real-time monitored environmental data is directly input into the model to dynamically adjust the solution process of the Navier-Stokes equation. In this way, the model can reflect the impact of environmental changes on spray diffusion in real time and enhance the accuracy and adaptability of the prediction.
[0129] The diffusion range prediction submodule also includes simulation of diffusion path and terminal distribution and spray range prediction. The simulation of diffusion path and terminal distribution is achieved by numerically solving the NavierStokes equation, simulating the movement path of spray particles from the nozzle until they reach the terminal distribution, and calculating the position and velocity of spray particles at different time and space points, thereby predicting the overall diffusion range of the spray. The spray range prediction determines the actual coverage area of the spray based on the simulation results.
[0130] The numerical solution of the Navier-Stokes equations is as follows:
[0131] Meshing: Divide the computational domain into grids or elements so that the equations can be solved in each small area;
[0132] Discretization equations: The continuous Navier-Stokes equations are discretized using the finite volume method and converted into a system of equations that can be solved on a computer. For each control volume, the Navier-Stokes equations are integrated to obtain:
[0133] ,in, represents the control volume, , , , represent velocity, pressure, density and kinematic viscosity respectively, represents the spray particle resistance, Represents a tiny volume element. It is used to represent an infinitesimal volume portion within the control volume during integration. When integrating fluid dynamics equations, it is necessary to integrate the fluid properties within the control volume (CV). Used to represent volume elements in these integrals;
[0134] Solving discrete equations: Apply iterative solvers to solve the discretized equations to obtain approximate solutions of flow field parameters such as velocity and pressure. Through the iterative process, the true values of flow field parameters such as velocity and pressure are gradually approached;
[0135] Particle tracking simulation: Use the obtained flow field data to simulate the movement path of the spray particles, and use the flow field data to perform particle tracking simulation, specifically: initialize the particle position and velocity; in each time step, according to the flow field velocity and external forces (such as gravity, wind force, etc.) at the current particle position, update the particle velocity and position:
[0136] ;
[0137] ;in, and are the velocity and position of the particle at time t, is the total force acting on the spray particles, is the particle mass, is the time step.
[0138] Spraying range forecast includes:
[0139] Path integration: By integrating all the particle paths obtained from the simulation, the overall diffusion model of the spray is established, which is achieved through the following steps:
[0140] Path data collection: During the particle tracking simulation, the position data of each particle is collected to form a path data set. , whose path consists of the time series of location points composition.
[0141] Path integration: The path data of all particles are integrated into a unified model by constructing a three-dimensional spatial distribution map, where the density value of each point represents the frequency of particle occurrence at that location.
[0142] Diffusion model construction: The diffusion model is generated using the path data, and voxelization technology (dividing the space into small cubes, each cube records the number of particles passing through the volume) is used to quantify the distribution of the spray in space;
[0143] Terminal distribution analysis: Evaluate the terminal distribution of particles in time and space to determine the coverage area and shape of the spray. The specific steps include:
[0144] Endpoint data collection: At the end of the simulation, collect the final positions of all particles .
[0145] Spatial distribution analysis: Analyze the spatial distribution of these end points by calculating the density of particles in different spatial regions. For example, the calculation area can be divided into grids, and the number of particles in each grid is calculated to obtain the spatial density distribution of the spray.
[0146] Coverage area and shape determination: Based on the spatial density distribution, the coverage area and shape of the spray are determined. A density threshold can be set to define the boundary of the spray range, that is, the area with density greater than a certain threshold is considered as the spray coverage area.
[0147] Through the above process, the spraying range can be accurately predicted, including the coverage area and shape of the spray. These calculations and analyses provide a scientific basis for accurately controlling the spraying pressure and optimizing the spraying range.
[0148] The pressure adjustment logic submodule specifically includes:
[0149] Determine the insufficient coverage range of spraying: Assume that the complete coverage range of adjacent sprinklers is , the actual coverage is ,if , then there is insufficient coverage;
[0150] Calculate the undercoverage ratio: Undercoverage ratio Calculated by the following formula:
[0151] ;
[0152] Relating spray pressure to spray range: Spray pressure And spraying range The linear relationship between them is expressed as: ,in, is the coefficient of relationship between spray range and pressure, is a constant;
[0153] Calculate pressure compensation: To achieve full coverage, determine the amount of pressure increase that will compensate for the lack of coverage , it is deduced that: ; Solve this equation to get the required pressure compensation value: ; or based on the undercoverage ratio To calculate: ;
[0154] Adjust the spray pressure: adjust the current spray pressure Add the calculated pressure compensation value , get the new spray pressure: ; Apply new spray pressure , adjust the spraying range to achieve ideal full coverage effect.
[0155] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.
[0156] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. An intelligent control system for spraying pressure of a spraying rod of a plant protection robot. It is characterized in that It includes an initial pressure setting module and a real-time pressure adjustment module; The initial pressure setting module sets the initial spraying pressure according to the distance between adjacent spray heads of the spraying rod, so that the spraying range between every two adjacent spray heads is fully covered; During the actual spraying operation, the real-time pressure adjustment module monitors the spraying coverage between two adjacent nozzles in real time to determine whether the coverage is complete. If the coverage is incomplete, the spraying pressure is adjusted; if the coverage is complete, no adjustment is made; the real-time pressure adjustment module also includes: Spray pattern capture submodule: uses a high-speed camera to capture the umbrella-shaped spray image formed by the pesticide sprayed from the nozzle; Morphological analysis submodule: Identify the boundary and structure of umbrella-shaped spray through edge detection and morphological analysis; Diffusion range prediction submodule: uses a kinetic model to simulate the diffusion path and terminal distribution of spray particles to obtain the spray range; Pressure adjustment logic submodule: Based on the spraying range, calculate the spraying pressure compensation value and adjust the spraying pressure to correct the deviation.
2. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 1, It is characterized in that The initial pressure setting module adjusts the spray pressure in the spray rod by operating the total pressure valve, observes the spraying range between each two adjacent nozzles, and continues to adjust the total pressure valve until it is confirmed that the spraying range of each adjacent nozzle is completely overlapped.
3. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 2, It is characterized in that The spray pattern capturing submodule uses at least one high-speed camera installed on the spraying rod, aimed at the spray outlet of at least one spray head, to capture the umbrella-shaped spray image formed by the sprayed pesticide in real time, and record the image of the pesticide from spraying to initial diffusion.
4. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 2, It is characterized in that The morphological analysis submodule specifically includes: Image preprocessing: The spray images captured by the high-speed camera were first preprocessed, including denoising, contrast enhancement, and color correction; Edge detection: Use edge detection algorithm to process the preprocessed image, identify the boundary line of the spray, reveal the geometric shape and external contour of the spray, and provide a basis for understanding the spray range; Morphological analysis: Based on edge detection, morphological analysis is performed, including erosion and dilation processing, to eliminate noise and fill in the discontinuities in the spray image; Spray feature extraction: Based on the results of edge detection and morphological analysis, key spray features are extracted, including the maximum diameter of the spray, the average diffusion angle, the particle size distribution and the diffusion velocity.
5. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 4, It is characterized in that The edge detection algorithm adopts the Canny edge detection algorithm, which is as follows: Gaussian filter: To adapt to the umbrella-shaped spray image, adjust the standard deviation of the Gaussian filter To capture the width and edge fluidity of the spray, it is calculated as: ,in, is the Gaussian filter for position Response, The value is selected based on the average size of the spray diameter to ensure that the spray edge remains clear after filtering; Gradient calculation and direction: Calculate the gradient strength and direction of the image to determine the boundary. The gradient strength is calculated as follows: ;in, and They are the gradients of the image in the x and y directions, and the gradient directions for: ; Non-maximum suppression: For spray images, when performing non-maximum suppression, the algorithm is adjusted to adapt to the curved shape of the spray to ensure that the edges are detected along the natural curve of the spray; Hysteresis threshold: Use two thresholds to binarize the gradient image and determine the true edge points. If the gradient value is higher than the high threshold, it is considered an edge point; if the gradient value is lower than the low threshold, it is excluded; if the gradient value is between the two thresholds, it is retained only if it is connected to the edge point.
6. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 5, It is characterized in that The morphological analysis: Morphological operations are used to analyze and process image structures, including corrosion and expansion: The erosion operation is used to remove details from the image boundary and is defined by the following formula: ,in, is the original image, is a structural element, Is a structural element Pan to Position the situation at that time; The dilation operation is used to enhance features in an image and is defined as: ,in, represents the intersection, which means that as long as and exist The locations overlap. belongs to the result image.
7. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 6, It is characterized in that In the non-maximum suppression, for each pixel, its neighboring pixels in the gradient direction are checked. If the intensity of the neighboring pixel in the gradient direction is not the largest, its intensity is set to 0, and only the pixel with the local maximum intensity is retained to achieve edge refinement. The specific steps are as follows: For each pixel in the image, according to its gradient direction , find two adjacent pixels in this direction; Compare the gradient strength of the current pixel With the gradient strength of two adjacent pixels, if the gradient strength of the current pixel is not the largest, the pixel is considered not to belong to the edge and its strength is set to 0; Repeat the above process until all pixels of the entire image have been processed.
8. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 1, It is characterized in that The diffusion range prediction submodule specifically includes: Receiving morphological analysis data: The diffusion range prediction submodule receives the spray image data from the morphological analysis submodule, including spray boundary structure, angle, diameter and particle distribution information; Dynamic model application: Based on the received data, a customized fluid dynamics model is used to simulate the movement and diffusion behavior of the spray particles in the air. Initial conditions are set in the customized fluid dynamics model. The boundary structure of the spray is used to set the initial velocity field and pressure field of the customized fluid dynamics model. The customized fluid dynamics model includes the Navier-Stokes equations, the spray particle behavior model, and the environmental condition interaction model. The NavierStokes equations are expressed as: ,in, is the velocity field, is the density, It's pressure. is the kinematic viscosity, Represents the resistance caused by the spray particle size and distribution, is the external force term, i.e. the influence of environmental wind on the spray; The spray particle behavior model considers the movement and interaction of a single spray particle or a particle group, and uses the following equation to represent the motion state of the particle: ; ; in, and are the position and velocity of the particle, is the relaxation time of the particle, which indicates the time required for the particle velocity to adjust to the airflow velocity, is the local air velocity, is the acceleration due to gravity; The environmental condition interaction model is used to describe the behavior of the spray under the influence of external conditions by changing the external force term in the Navier-Stokes equation. To simulate the effects of wind: ,in, is the air density, is the drag coefficient related to particle shape and size, is the wind speed vector, is the velocity vector of the particle or fluid; The diffusion range prediction submodule also includes simulating the diffusion path and terminal distribution and spraying range prediction, wherein the simulated diffusion path and terminal distribution are simulated by numerically solving the NavierStokes equation, simulating the movement path of the spray particles starting from the nozzle until reaching the terminal distribution, and calculating the position and speed of the spray particles at different time and space points, thereby predicting the overall diffusion range of the spray; the spraying range prediction determines the actual coverage area of the spray based on the simulation results.
9. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 8, It is characterized in that The numerical solution of the Navier-Stokes equation is as follows: Meshing: Divide the computational domain into grids or elements so that the equations can be solved in each small area; Discretization equations: The continuous Navier-Stokes equations are discretized using the finite volume method and converted into a system of equations that can be solved on a computer. For each control volume, the Navier-Stokes equations are integrated to obtain: ,in, represents the control volume, , , , represent velocity, pressure, density and kinematic viscosity respectively, represents the spray particle resistance, Represents a tiny volume element; Solving Discrete Equations: Applying iterative solvers to solve the discretized system of equations, obtaining approximate solutions for flow field parameters such as velocity and pressure, through an iterative process; Particle tracking simulation: Use the obtained flow field data to simulate the motion path of the spray particles, and use the flow field data to perform particle tracking simulation, specifically: initialize the particle position and velocity; in each time step, update the particle velocity and position according to the flow field velocity and external force at the current particle position: ; ;in, and are the velocity and position of the particle at time t, is the total force acting on the spray particles, is the particle mass, is the time step; The spraying range prediction specifically includes: Path integration, by integrating all the particle paths obtained by simulation, an overall diffusion model of the spray is established; Terminal distribution analysis evaluates the terminal distribution of particles in time and space to determine the coverage area and shape of the spray.
10. The intelligent control system for spraying pressure of a spraying rod of a plant protection robot according to claim 9, It is characterized in that The pressure adjustment logic submodule specifically includes: Determine the insufficient coverage range of spraying: Assume that the complete coverage range of adjacent sprinklers is , the actual coverage is ,if , then there is insufficient coverage; Calculate the undercoverage ratio: Undercoverage ratio Calculated by the following formula: ; Relating spray pressure to spray range: Spray pressure And spraying range The linear relationship between them is expressed as: ,in, is the coefficient of relationship between spray range and pressure, is a constant; Calculate pressure compensation: To achieve full coverage, determine the amount of pressure increase that will compensate for the lack of coverage , it is deduced that: ; Solve this equation to get the required pressure compensation value: ; or based on the undercoverage ratio To calculate: ; Adjust the spray pressure: adjust the current spray pressure Add the calculated pressure compensation value , get the new spray pressure: ; Apply new spray pressure , adjust the spraying range to achieve ideal full coverage effect.
