A tea leaf picking robot picking path generation method and device

Abstract

The application provides a tea leaf picking robot picking path generation method and device, wherein the method comprises obtaining tea bud leaf information based on a vision system of a tea leaf picking robot, and generating a tea bud leaf three-dimensional envelope based on the tea bud leaf information. A multi-objective optimization model is constructed, and the optimal picking path of the mechanical arm under the multi-objective is calculated according to the multi-objective optimization model. The optimal picking path of the mechanical arm is curve-fitted to obtain the picking path of the tea leaf picking robot. The tea bud leaf information comprises tea bud leaf picking point position coordinates and the three-dimensional space posture of the tea bud leaf. The method provided by the application can utilize the three-dimensional space coordinates and posture of the tea bud leaf to automatically generate the picking path of the tea leaf picking robot according to the three-dimensional space coordinates and posture, thereby improving the tea bud leaf picking success rate, efficiency and endurance of the tea leaf picking robot.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing and path generation, in particular to a picking path generation method and device of a tea picking robot. BACKGROUND

[0002] Tea is rich in tea polyphenols, amino acids, vitamins and other beneficial health bioactive ingredients and nutrients, and its unique pharmacological function makes it popular among consumers. However, the production cost of tea, especially the picking cost of famous tea, has been rising, which has become a key factor restricting the industry's efficiency. The picking of famous tea highly depends on manual labor, and the workload of famous tea picking can account for more than 70% of the entire production process. Not only is it the most labor-intensive link, but also the most costly part. This link has significant seasonal and time constraints, with short operation cycles and intensive tasks, requiring a large amount of manpower for repetitive labor within a limited time.

[0003] For a famous tea picking robot system, the motion trajectory planning of the mechanical arm is one of the core technical problems of its control system. Efficient and reasonable picking path planning directly determines the operation efficiency and picking effect of the robot. In view of the current development needs of automatic picking machinery for famous tea, the technical problem to be solved is how to design and optimize the picking sequence, generate the motion trajectory of the mechanical arm according to the picking sequence, so as to realize efficient and energy-saving and low-physical-damage picking operation.

[0004] Therefore, a method is needed that can use the three-dimensional spatial coordinates and attitude of tea buds and leaves to automatically generate the picking path of a tea picking robot, thereby improving the success rate and efficiency of tea bud and leaf picking of the tea picking robot. SUMMARY

[0005] To overcome the problems in the related art, the purpose of the present application is to provide a picking path generation method and device of a tea picking robot, which can use the three-dimensional spatial coordinates and attitude of tea buds and leaves to automatically generate the picking path of a tea picking robot, thereby improving the success rate and efficiency of tea bud and leaf picking of the tea picking robot.

[0006] A picking path generation method of a tea picking robot, comprising:

[0007] Obtaining tea bud and leaf information based on the vision system of the tea picking robot;

[0008] Generating a three-dimensional envelope of tea buds and leaves based on the tea bud and leaf information;

[0009] Constructing a multi-objective optimization model;

[0010] Calculating the optimal picking path of the mechanical arm under multiple objectives according to the multi-objective optimization model;

[0011] Curve fitting is performed on the optimal picking path of the mechanical arm to obtain a picking path of the tea picking robot.

[0012] In the preferred technical solution of the present application, the tea bud leaf information is acquired based on a vision system of the tea picking robot, and the tea bud leaf information includes:

[0013] The petiole axis direction vector of the tea bud leaf in a three-dimensional space is calculated using the vision system of the tea picking robot, and the petiole axis direction vector of the tea bud leaf is used to reflect the three-dimensional space posture of the tea bud leaf.

[0014] According to the spatial geometric relationship between the petiole axis direction vector of the tea bud leaf and the leaf growth direction of the tea bud leaf, the position coordinates of the tea bud leaf picking point are determined.

[0015] In the preferred technical solution of the present application, the tea bud leaf three-dimensional envelope is generated based on the tea bud leaf information, and the tea bud leaf three-dimensional envelope includes:

[0016] The petiole axis direction vector of the tea bud leaf is taken as the cylindrical central axis direction vector.

[0017] The position coordinates of the tea bud leaf picking point are taken as the bottom surface center point coordinates.

[0018] The cylindrical radius and the cylindrical height are acquired, and the tea bud leaf three-dimensional envelope is drawn according to the cylindrical central axis direction vector, the bottom surface center point coordinates, the cylindrical radius and the cylindrical height.

[0019] In the preferred technical solution of the present application, the multi-objective optimization model is constructed, and the multi-objective optimization model includes:

[0020] The path length function of the end effector of the tea picking robot is constructed in combination with the mechanical arm constraint condition.

[0021] The collision risk function of the end effector of the tea picking robot is constructed.

[0022] The energy consumption function of the end effector of the tea picking robot is constructed.

[0023] The optimal solution set under multi-objective is obtained by optimization in combination with the path length function, the minimum collision risk function and the energy consumption function.

[0024] In the preferred technical solution of the present application, the path length function of the end effector of the tea picking robot includes:

[0025] According to the following formula, the path length function is calculated based on the polygon motion rule of the tea bud leaf in a three-dimensional space.

[0026]

[0027] Wherein, n is the number of tea bud leaves, P end,i is the three-dimensional coordinates of the i-th preparatory picking point along the petiole axis direction, P end,i+1 is the three-dimensional coordinates of the i+1-th preparatory picking point along the petiole axis direction, P pick,i is the three-dimensional coordinates of the i-th tea bud leaf picking point; O n is the starting coordinates of the mechanical arm, P end,1 is the three-dimensional coordinates of the 1st preparatory picking point along the petiole axis direction, D is the path length function.

[0028] In the preferred technical solution of the present application, the collision risk function of the end effector of the tea leaf picking robot comprises:

[0029] The collision risk function is calculated according to the following formula:

[0030]

[0031] Wherein, V i is the three-dimensional envelope volume of the i-th tea bud leaf, V j is the three-dimensional envelope volume of the j-th tea bud leaf, is an indicator function, n is the number of tea bud leaves, R is the collision risk function; if the collision risk function is 1, the tea bud leaf with the largest z-axis coordinate is selected.

[0032] In the preferred technical solution of the present application, the energy consumption function of the end effector of the tea leaf picking robot comprises:

[0033] Based on the minimum energy consumption rule of each joint of the mechanical arm, the energy consumption function is calculated according to the following formula:

[0034]

[0035] Wherein, m is the number of joints of the mechanical arm, θ i,j is the angle value of the j-th joint of the mechanical arm at the i-th point, θ i+1,j is the angle value of the j-th joint of the mechanical arm at the i+1-th point, θ n,j is the angle value of the j-th joint of the mechanical arm at the n-th point, θ 0,j is the angle value of the j-th joint of the mechanical arm at the initial position, E is the energy consumption function.

[0036] In the preferred technical solution of the present application, the path length function, the minimum collision risk function and the energy consumption function are combined for optimization to obtain an optimal solution set under multi-objective optimization, comprising:

[0037] The multi-objective optimization model is optimized according to the following formula:

[0038]

[0039] F(x)=[D,R,E] T ;

[0040] wherein, is a kinematic constraint of the robot arm, min is a minimum function, F(x) is a function of the multi-objective optimization model, Z is a minimum value of the function of the multi-objective optimization model, D is a path length function, R is a collision risk function, E is an energy consumption function, x is a selected path, and T is a matrix transpose.

[0041] In the preferred technical solution of the present application, the optimal picking path of the robot arm is curve-fitted, comprising:

[0042] A cubic B-spline curve is selected according to the following formula to fit the expected path and obtain a motion trajectory curve:

[0043]

[0044] wherein, n+1 is the number of control points, N i,3 (t) is a cubic B-spline basis function, P i is the i-th control point, S(t) is the motion trajectory curve, t is a node vector, t min is the minimum node vector, and t max is the maximum node vector.

[0045] A picking path generation device of a tea leaf picking robot, comprising:

[0046] A tea bud leaf information acquisition module is configured to acquire tea bud leaf information based on a vision system of the tea leaf picking robot.

[0047] A tea bud leaf three-dimensional envelope generation module is configured to generate a tea bud leaf three-dimensional envelope based on the tea bud leaf information.

[0048] A model construction module is configured to construct a multi-objective optimization model.

[0049] A path calculation module is configured to calculate an optimal picking path of a robot arm under a multi-objective based on the multi-objective optimization model.

[0050] A path fitting module is configured to curve-fit the optimal picking path of the robot arm to obtain a picking path of the tea leaf picking robot.

[0051] The present application has the following advantages:

[0052] The tea-picking robot path generation method provided by this invention includes acquiring tea bud and leaf information based on the robot's vision system. This information includes the coordinates of the picking point and the three-dimensional spatial posture of the tea buds and leaves. A three-dimensional envelope of the tea buds and leaves is generated based on the cylinder's central axis direction vector, the coordinates of the base center point, the cylinder radius, and the cylinder height. A multi-objective optimization model is constructed, establishing the path length function, collision risk function, and energy consumption function of the robot's end effector. The multi-objective optimization model is optimized based on these functions to obtain the optimal solution set under multiple objectives. The optimal picking path for the robotic arm under multiple objectives is calculated based on this optimal solution set. Curve fitting is then performed on the optimal picking path to obtain the tea-picking robot's picking path. This method can automatically generate the tea-picking robot's picking path using the three-dimensional spatial coordinates and posture of the tea buds and leaves, thereby improving the success rate and efficiency of tea bud and leaf picking. Attached Figure Description

[0053] Figure 1 This is a flowchart of the tea-picking robot's path generation method according to the present invention;

[0054] Figure 2 This is a flowchart of the present invention for generating a three-dimensional envelope of tea buds and leaves based on tea bud and leaf information. Detailed Implementation

[0055] Preferred embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0056] Example 1

[0057] like Figure 1 As shown, this embodiment provides a method for generating the picking path of a tea-picking robot, including:

[0058] S1: Acquire information about tea buds and leaves using a vision system based on a tea-picking robot.

[0059] S2: Generate a three-dimensional envelope of tea buds and leaves based on the tea bud and leaf information.

[0060] S3: Construct a multi-objective optimization model.

[0061] S4: Calculate the optimal picking path of the robotic arm under multiple objectives based on the multi-objective optimization model.

[0062] S5: Curve fitting is performed on the optimal picking path of the mechanical arm to obtain a picking path of the tea leaf picking robot.

[0063] The tea leaf picking robot based vision system acquires tea bud leaf information, including:

[0064] S11: The tea leaf picking robot vision system is used to calculate the petiole axis direction vector of the tea bud leaf in three-dimensional space, which reflects the three-dimensional space posture of the tea bud leaf.

[0065] S12: According to the spatial geometric relationship between the petiole axis direction vector of the tea bud leaf and the leaf growth direction of the tea bud leaf, the position coordinates of the tea bud leaf picking point are determined.

[0066] The tea leaf picking robot has a vision system, a power system, a picking system, and a signal processing system. The picking system includes a mechanical arm, the power system includes a battery and a connecting shaft, and the vision system includes a camera. The battery provides power for other components of the tea leaf picking robot, the camera captures tea bud leaf images in the tea garden scene, and the signal processing system receives the tea bud leaf images and executes steps S1-S5 to generate the picking path of the tea leaf picking robot.

[0067] The tea bud leaf is a leaf bud differentiated from the top of the tea tree branch, and the petiole is the part connecting the leaf blade and the stem of the tea bud leaf. The petiole plays a role in supporting the leaf blade, transporting nutrients such as organic matter and inorganic elements, and transporting water. In the leaf structure of tea trees, the length and thickness of the petiole are affected by factors such as variety and growing environment.

[0068] The vision system captures tea bud leaf images in the tea garden, and the tea bud leaf images are three-dimensional images. The signal processing system sequentially performs gray scale conversion, image filtering, and edge detection on the tea bud leaf images to extract the petiole base point and the petiole end point of the tea bud leaf. The petiole base point is the point where the petiole connects with the tea stem, and the petiole end point is the point where the petiole connects with the leaf blade.

[0069] The three-dimensional coordinates of the petiole end point are subtracted from the three-dimensional coordinates of the petiole base point to obtain the petiole axis direction vector of the tea bud leaf in three-dimensional space, i.e. the growth main axis of the tea bud leaf. The petiole axis direction vector of the tea bud leaf can reflect the three-dimensional space posture of the tea bud leaf. After determining the petiole axis direction vector of the tea bud leaf, the intersection between the petiole axis direction vector of the tea bud leaf and the leaf growth direction of the tea bud leaf is detected to obtain the position coordinates of the tea bud leaf picking point.

[0070] As shown in Figure 2 The tea bud leaf three-dimensional envelope is generated based on the tea bud leaf information, including:

[0071] S21: The petiole axis direction vector of the tea bud leaf is used as the cylindrical central axis direction vector.

[0072] S22: Take the position coordinates of the tea bud leaf picking point as the bottom center point coordinates.

[0073] S23: Obtain the cylinder radius and the cylinder height, and draw the tea bud leaf three-dimensional envelope according to the cylinder central axis direction vector, the bottom center point coordinates, the cylinder radius and the cylinder height.

[0074] The tea bud leaf three-dimensional envelope is cylindrical, the direction of the tea bud leaf three-dimensional envelope is determined according to the petiole axis direction vector of the tea bud leaf, and the position of the bottom surface of the tea bud leaf three-dimensional envelope is determined according to the bottom center point coordinates. The size of the tea bud leaf three-dimensional envelope is determined according to the cylinder radius and the cylinder height, and the tea bud leaf three-dimensional envelope is drawn in combination with the cylinder central axis direction vector, the bottom center point coordinates, the cylinder radius and the cylinder height.

[0075] The curve fitting of the optimal picking path of the mechanical arm includes:

[0076] The cubic B-spline curve is selected to fit the expected path according to the following formula to obtain the motion trajectory curve:

[0077]

[0078] Where n+1 is the number of control points, N i,3 (t) is the cubic B-spline basis function, P i is the i-th control point, S(t) is the motion trajectory curve, t is the node vector, t min is the minimum node vector, t max is the maximum node vector.

[0079] The cubic B-spline curve has good local controllability, shape preservation and numerical stability, etc. The cubic B-spline curve is defined by a group of control points and a node vector. The cubic B-spline curve is selected to fit the expected path to obtain a smooth and continuous motion trajectory curve.

[0080] The tea leaf picking robot picking path generation method provided by the embodiment comprises: acquiring tea bud leaf information based on a vision system of a tea leaf picking robot, the tea bud leaf information comprising tea bud leaf picking point position coordinates and a three-dimensional space posture of the tea bud leaf; generating a tea bud leaf three-dimensional envelope according to a cylinder central axis direction vector, a bottom surface center point coordinate, a cylinder radius and a cylinder height; constructing a multi-objective optimization model; establishing a path length function, a collision risk function and an energy consumption function of an end effector of the tea leaf picking robot; optimizing the multi-objective optimization model according to the path length function, the collision risk function and the energy consumption function to obtain a multi-objective optimal solution set; calculating a mechanical arm optimal picking path under the multi-objective according to the multi-objective optimal solution set of the multi-objective optimization model; and performing curve fitting on the mechanical arm optimal picking path to obtain a picking path of the tea leaf picking robot. The method provided by the embodiment can automatically generate a picking path of a tea leaf picking robot according to three-dimensional space coordinates and a posture of a tea bud leaf, thereby improving a tea bud leaf picking success rate and efficiency of the tea leaf picking robot.

[0081] Embodiment 2

[0082] As Figure 1 shown, the embodiment provides a tea leaf picking robot picking path generation method, and the embodiment describes differences from embodiment 1 on the basis of embodiment 1. The method comprises:

[0083] S1: acquiring tea bud leaf information based on a vision system of a tea leaf picking robot.

[0084] S2: generating a tea bud leaf three-dimensional envelope based on the tea bud leaf information.

[0085] S3: constructing a multi-objective optimization model.

[0086] S4: calculating a mechanical arm optimal picking path under the multi-objective according to the multi-objective optimization model.

[0087] S5: performing curve fitting on the mechanical arm optimal picking path to obtain a picking path of the tea leaf picking robot.

[0088] The multi-objective optimization model is constructed, comprising:

[0089] S31: constructing a path length function of an end effector of the tea leaf picking robot in combination with a mechanical arm constraint condition.

[0090] S32: constructing a collision risk function of the end effector of the tea leaf picking robot.

[0091] S33: constructing an energy consumption function of the end effector of the tea leaf picking robot.

[0092] S34: optimization is performed in combination with the path length function, the minimum collision risk function and the energy consumption function to obtain an optimal solution set under multi-objective.

[0093] The path length function of the end effector of the tea leaf picking robot comprises:

[0094] The path length function is calculated according to the following formula based on the polygon motion rule of tea buds in three-dimensional space:

[0095]

[0096] Wherein, n is the number of tea buds, P end,i is the three-dimensional coordinate of the i-th preliminary picking point along the petiole axis direction, P end,i+1 is the three-dimensional coordinate of the i+1-th preliminary picking point along the petiole axis direction, P pick,i is the three-dimensional coordinate of the i-th tea bud picking point; O n is the starting coordinate of the mechanical arm, P end,1 is the three-dimensional coordinate of the 1st preliminary picking point along the petiole axis direction, and D is the path length function.

[0097] The three-stage disturbance-free polygon motion rule of lifting-translation-descending based on the three-dimensional space posture of tea buds is constructed, and the two-norm of the difference between P end,i and P pick,i is used to measure the difference between the three-dimensional coordinates of the i-th preliminary picking point and the i-th tea bud picking point. The two-norm of the difference between P end,i+1 and P end,i is used to measure the difference between adjacent preliminary picking points. The two-norm of the difference between O n and P end,1 is used to measure the difference between the three-dimensional coordinates of the starting coordinate of the mechanical arm and the 1st preliminary picking point.

[0098] The collision risk function of the end effector of the tea leaf picking robot comprises:

[0099] The collision risk function is calculated according to the following formula:

[0100]

[0101] Wherein, V i is the three-dimensional envelope volume of the i-th tea bud, V j is the three-dimensional envelope volume of the j-th tea bud, is an indicator function, n is the number of tea buds, and R is the collision risk function; if the collision risk function is 1, the tea bud with the largest z-axis coordinate is selected.

[0102] When the three-dimensional envelope volumes Vi and V j When the two tea buds interfere with each other, that is, there is an overlapping area, the indicator function is 1, at this time, the upper tea leaf is preferentially selected, that is, the tea bud leaf with a larger z-axis coordinate is selected. When the three-dimensional envelope bodies V i and V j When the two tea buds do not interfere with each other, that is, there is no overlapping area, the indicator function is 0.

[0103] The energy consumption function of the end effector of the tea leaf picking robot is constructed, and the energy consumption function includes:

[0104] Based on the rule of minimum energy consumption of each joint of the mechanical arm, the energy consumption function is calculated according to the following formula:

[0105]

[0106] Wherein, m is the number of joints of the mechanical arm, θ i,j is the angle value of the jth joint of the mechanical arm at the ith point, θ i+1,j is the angle value of the jth joint of the mechanical arm at the i+1th point, θ n,j is the angle value of the jth joint of the mechanical arm at the nth point, θ 0,j is the angle value of the jth joint of the mechanical arm at the initial position, and E is the energy consumption function.

[0107] A norm is used to measure the difference between the angle values of the joints of the same serial number of adjacent points, and a norm is used to measure the difference between the angle values of the jth joint at the nth point and the jth joint at the initial position.

[0108] The path length function, the minimum collision risk function and the energy consumption function are combined for optimization to obtain an optimal solution set under multi-objective optimization, which includes:

[0109] The multi-objective optimization model is optimized according to the following formula:

[0110]

[0111] F(x)=[D,R,E] T ;

[0112] Wherein, is the kinematic constraint of the mechanical arm, min is the minimum function, F(x) is the function of the multi-objective optimization model, Z is the minimum value of the function of the multi-objective optimization model, D is the path length function, R is the collision risk function, E is the energy consumption function, x is the selected path, and T is the matrix transpose.

[0113] The path length target function D, the collision risk function R and the energy consumption model V are regarded as three independent optimization targets in the embodiment, and the Pareto front selection is realized through non-dominated sorting and congestion calculation based on a multi-objective optimization algorithm. The specific steps are to use the NSGA-II algorithm to perform initialization of the population, non-dominated sorting, congestion calculation, crossover mutation and environment selection, and finally to generate the Pareto front. The Pareto front refers to the surface formed by all optimal sets in space, and the path with the minimum priority, the lowest collision risk and the minimum energy consumption is selected. At the same time, the path length function, the collision risk function and the energy consumption function are optimized, which can minimize the motion path of the robot arm, reduce the collision between the robot arm and the tea buds in the process of picking tea buds, and reduce the motion energy consumption of each joint of the robot arm.

[0114] Embodiment 3

[0115] The embodiment provides a picking path generation device of a tea leaf picking robot, comprising:

[0116] a tea bud information acquisition module configured to acquire tea bud information based on a vision system of the tea leaf picking robot;

[0117] a tea bud three-dimensional envelope generation module configured to generate a tea bud three-dimensional envelope based on the tea bud information;

[0118] a model construction module configured to construct a multi-objective optimization model;

[0119] a path calculation module configured to calculate an optimal picking path of a robot arm under the multi-objective optimization model;

[0120] a path fitting module configured to perform curve fitting on the optimal picking path of the robot arm to obtain a picking path of the tea leaf picking robot.

[0121] The picking path generation device of the tea leaf picking robot in the embodiment is used to execute the picking path generation method of the tea leaf picking robot in Embodiments 1 and 2.

[0122] It should be noted that in this document, the terms "comprising", "containing" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, device, article or method. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of another identical element in the process, device, article or method including the element.

[0123] The above merely describes the preferred embodiments of the present application, and is not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, which is made according to the content of the present application specification and drawings, is also included in the patent protection scope of the present application.

Claims

1. A method for generating a picking path of a tea leaf picking robot, characterized by, The method comprises the following steps: acquiring tea bud leaf information based on a vision system of a tea leaf picking robot; generating a three-dimensional envelope of the tea bud leaf based on the tea bud leaf information; constructing a multi-objective optimization model; calculating an optimal picking path of a mechanical arm under multi-objective conditions according to the multi-objective optimization model; curve fitting the optimal picking path of the mechanical arm to obtain a picking path of the tea leaf picking robot; the construction of the multi-objective optimization model comprises: constructing a path length function of an end effector of the tea leaf picking robot in combination with a mechanical arm constraint condition; constructing a collision risk function of the end effector of the tea leaf picking robot; constructing an energy consumption function of the end effector of the tea leaf picking robot; optimizing the path length function, the minimum collision risk function and the energy consumption function to obtain an optimal solution set under multi-objective conditions; the construction of the path length function of the end effector of the tea leaf picking robot comprises: calculating the path length function according to the following formula based on a polygon motion rule of the tea bud leaf in a three-dimensional space: ； wherein n is the number of tea bud leaves, P end,i is the three-dimensional coordinate of the i-th preparatory picking point along the axis direction of the petiole, P end,i+1 is the three-dimensional coordinate of the i+1-th preparatory picking point along the axis direction of the petiole, P pick,i is the three-dimensional coordinate of the i-th tea bud leaf picking point; O n is the starting coordinate of the mechanical arm, P end,1 is the three-dimensional coordinate of the 1st preparatory picking point along the axis direction of the petiole, D is a path length function; the construction of the collision risk function of the end effector of the tea leaf picking robot comprises: calculating the collision risk function according to the following formula: ； wherein V i is the three-dimensional envelope volume of the i-th tea bud leaf, V j is the three-dimensional envelope volume of the j-th tea bud leaf, V is an indicator function, n is the number of tea bud leaves, and R is a collision risk function; if the collision risk function is 1, the tea bud leaf with the largest z-axis coordinate is selected. the construction of the energy consumption function of the end effector of the tea leaf picking robot comprises: calculating the energy consumption function according to the following formula based on a minimum energy consumption rule of each joint of the mechanical arm: ； wherein m is the number of joints of the robot arm, is the angle value of the jth joint of the robot arm at the ith point, is the angle value of the jth joint of the robot arm at the ith+1 point, is the angle value of the jth joint of the robot arm at the nth point, is the angle value of the jth joint of the robot arm at the initial position, and E is an energy consumption function.

2. The picking path generation method of the tea leaf picking robot according to claim 1, characterized in that, the acquisition of the tea bud leaf information based on the vision system of the tea leaf picking robot comprises: calculating a petiole axis direction vector of the tea bud leaf in a three-dimensional space using the vision system of the tea leaf picking robot, the petiole axis direction vector of the tea bud leaf being used to reflect a three-dimensional space posture of the tea bud leaf; determining a position coordinate of a tea bud leaf picking point according to a space geometric relationship between the petiole axis direction vector of the tea bud leaf and a leaf blade growth direction of the tea bud leaf.

3. The picking path generation method of the tea leaf picking robot according to claim 2, characterized in that, the generation of the three-dimensional envelope of the tea bud leaf based on the tea bud leaf information comprises: taking the petiole axis direction vector of the tea bud leaf as a cylindrical central axis direction vector; taking the position coordinate of the tea bud leaf picking point as a bottom surface center point coordinate; obtaining a cylindrical radius and a cylindrical height, and drawing the three-dimensional envelope of the tea bud leaf according to the cylindrical central axis direction vector, the bottom surface center point coordinate, the cylindrical radius and the cylindrical height.

4. The picking path generation method of the tea leaf picking robot according to claim 1, characterized in that, the optimization of the path length function, the minimum collision risk function and the energy consumption function to obtain the optimal solution set under multi-objective conditions comprises: optimizing the multi-objective optimization model according to the following formula: ； ； wherein, is a kinematic constraint of the robot arm, min is a minimum function, F(x) is a function of the multi-objective optimization model, Z is a minimum value of the function of the multi-objective optimization model, D is a path length function, R is a collision risk function, E is an energy consumption function, x is a selected path, and T is a matrix transpose. 5.The picking path generation method of the tea leaf picking robot according to claim 1, characterized in that, the curve fitting of the optimal picking path of the mechanical arm comprises: selecting a cubic B-spline curve to fit the expected path according to the following formula to obtain a motion trajectory curve: ； where n+1 is the number of control points, N i,3 (t) is a cubic B-spline basis function, P i is the i-th control point, S(t) is the motion trajectory curve, t is the node vector, t min is the minimum node vector, t max is the maximum node vector.

6. A tea leaf picking robot's picking path generating device characterized by comprising: The device is used for executing the picking path generation method of the tea leaf picking robot in claim 1, and the device comprises: a tea bud leaf information acquisition module configured to acquire tea bud leaf information based on a vision system of a tea leaf picking robot; a tea bud leaf three-dimensional envelope generation module configured to generate a three-dimensional envelope of the tea bud leaf based on the tea bud leaf information; a model construction module configured to construct a multi-objective optimization model; a path calculation module configured to calculate an optimal picking path of a mechanical arm under multi-objective conditions according to the multi-objective optimization model; a path fitting module configured to curve fit the optimal picking path of the mechanical arm to obtain a picking path of the tea leaf picking robot.

Images