Parameter optimization design method and device for 2-RPR series robot

Abstract

The invention belongs to the related technical field of robot structure design, and discloses a parameter optimization design method and device for a 2-RPR series robot, and the method comprises the steps: (1) sequentially connecting coordinate points at two joints of the 2-RPR series robot, a lower hinge point of an electric cylinder, and upper hinge points of the electric cylinder corresponding to two limit positions to obtain an obstacle space of the robot; the obstacle space of the robot is an electric cylinder movement area which can be reached by the tail end of the robot theoretically and is not allowed to reach in actual use; (2) judging whether the working space of the robot envelops the target working space of the robot or not, further judging whether the target working space of the robot coincides with the obstacle space of the robot or not, and determining a coefficient k according to two judgment results; and (3) optimizing connecting rod parameters and branch chain parameters of the 2-RPR series robot by adopting a genetic algorithm. On the premise that the working space meets the requirement, the robot design parameters with the better connecting rod length and the better branch chain arrangement are obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field related to robot structure design, and more specifically, relates to a parameter optimization design method and equipment for a 2-RPR serial robot. Background Art

[0002] The size of the robot's workspace is mainly determined by the robot's work requirements. For serial robots, when the workspace meets the actual requirements, optimizing the robot parameters can improve the robot's spatial accessibility, make the robot structure more compact, reduce the robot's weight, and reduce manufacturing costs. The optimized workspace has a higher utilization rate and is more in line with actual work requirements.

[0003] In 2024, San Hongjun et al. optimized the parameters of the Delta parallel robot. By establishing an envelope penalty function, the kinematic performance evaluation function was obtained by using multivariate nonlinear fitting and linear weighted combination method. The optimization model was established by combining the condition number distribution characteristics and distortion constraints, and the optimization was performed using genetic algorithm. Compared with before optimization, the volume of the workspace can be reduced by 14.26% after optimization. In 2024, Xu Chenyu et al. used the WOA algorithm to comprehensively optimize the parameters and dimensions of the new 4RPUR parallel robot, making the workspace boundary smooth and symmetrically distributed. In 2024, Li Zhen et al. used the particle swarm algorithm to optimize the main structural parameters of the parallel machine leg. The optimized workspace volume of the mechanical leg increased by 69.83%. In 2014, Gan Yi et al. used genetic algorithm to optimize the DH parameters of the 6R robot under the given workspace, and obtained the optimal solution that meets the constraints and minimizes the robot's connecting rod length. In 2021, Hu Wangning used genetic algorithms to optimize the arm length of the PUMA arc welding robot, thereby optimizing the robot's working space, with an optimization rate of 5.5%.

[0004] Scholars have conducted a lot of optimization research on the structural parameters of parallel robots, expanding or reducing the robot's workspace. The optimized robot configuration can better serve the application scenarios. With the rapid development of marine equipment, the demand for efficient and high-quality processing of large propellers is increasing. The 2-RPR serial robot has become an effective solution for efficient and high-quality robot milling of large propellers with the advantage of high rigidity of the main body. Since this configuration consists of a six-axis series main body and a branched chain structure with dual electric cylinders, the connecting rod parameters and branch chain parameters are often redundant. This configuration also has a unique obstacle space caused by the movement of the electric cylinder, which will limit the robot's workspace. Summary of the invention

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a parameter optimization design method and device for a 2-RPR serial robot, which aims to solve the problem of redundancy in the connecting rod parameters and branch chain parameters of the 2-RPR serial robot.

[0006] To achieve the above object, according to one aspect of the present invention, a parameter optimization design method for a 2-RPR serial robot is provided, the method comprising the following steps:

[0007] (1) Connect the coordinate points of the two joints of the 2-RPR serial robot, the lower hinge point of the electric cylinder, and the upper hinge points of the electric cylinder corresponding to the two extreme positions in sequence to obtain the robot's obstacle space; the robot's obstacle space is the electric cylinder movement area that the robot end can theoretically reach but is not allowed to reach in actual use;

[0008] (2) Determine whether the robot's workspace envelops the robot's target workspace, and then determine whether the robot's target workspace overlaps with the robot's obstacle space, and determine the coefficient k based on the two judgment results;

[0009] (3) A genetic algorithm is used to optimize the connecting rod parameters and branch chain parameters of the 2-RPR serial robot; wherein the fitness function of the genetic algorithm is:

[0010] f i '=max{f1,f2,…,f n}-f i +rand()i=1,2,…,j

[0011] In the formula, j is the population size, f i =k*(ω1a2+ω2a3+ω3d4+ω4c1+ω5e1), ω1, ω2, ω3, ω4, ω5 are the weights of the five optimization variables a2, a3, d4, c1, e1 corresponding to population i respectively.

[0012] Furthermore, if the robot's workspace envelops the robot's target workspace, and the robot's target workspace does not overlap with the robot's obstacle space, the determination coefficient k is 1; otherwise, the determination coefficient k is 1.5.

[0013] Furthermore, when judging whether the robot's workspace envelops the robot's target workspace, three planes Y=0, Z=d1 and Z=d2 in the robot's workspace may be intercepted for judgment.

[0014] Further, for the Y=0 section: starting from the rightmost point of the section, i.e. the point with the largest x coordinate X max , record the coordinates of the point as (x max,z0), then when the z coordinate is in the range (z0-H / 2,z0+H / 2) and the x coordinate is less than x max -L / 2, find the point with the largest x coordinate among all the points in the workspace within x0+L / 2, and record the x coordinate value x0 of this point. At this time, the optimal position of the target workspace in this section is determined. The coordinates of the four vertices are A(x0,z0+H / 2), B(x0,z0-H / 2), C(x0+L,z0-H / 2), and D(x0+L,z0+H / 2). Since the values ​​of the abscissas of points AB are the abscissas of the concave points on the left contour of the workspace, the line segment formed by AB must be included in the robot workspace, and points CD need to be judged to find out whether the ordinates of the points in the workspace with abscissas x0+L / 2 are both less than the ordinate of point C and greater than the ordinate of point D. If there are points that meet the requirements, it means that the line segment formed by CD must also be included in the robot workspace, which means that the rectangle formed by points ABCD is included in the robot workspace.

[0015] Furthermore, according to the shape characteristics of the cross-sectional vertex coordinates of the machine obstacle space and the vertex coordinates of the target workspace, it is only necessary to determine whether point B in the target workspace is within the obstacle space at the Y=0 cross section.

[0016] Furthermore, the transformation matrices of each link of the 2-RPR serial robot are sequentially multiplied to obtain the relative transformation matrix of the end link coordinate system n relative to the base coordinate system 0, and then the random joint variables are substituted into the obtained relative transformation matrix to obtain the robot's workspace.

[0017] Furthermore, according to the configuration of the 2-RPR serial robot, the coordinate system of each link of the robot is established, and the link parameters are determined according to the coordinate system of each link. Then, the transformation matrices of each link are multiplied sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the polar coordinate system 0. The random joint variables generated by the random function rand() are substituted into the obtained relative transformation matrix to obtain the coordinate set {P|P(x, y, z)} of the end of the robot, and then the robot's workspace is obtained.

[0018] Furthermore, the random function rand() is used to generate random joint variables:

[0019]

[0020] Where q is the joint random variable; q max and q min are the upper and lower limits of the joint angle respectively; i=1,…,n, where n is the number of robot joints.

[0021] The present invention also provides a parameter optimization design system for a 2-RPR serial robot, the system comprising a memory and a processor, the memory storing a computer program, and the processor executing the parameter optimization design method for the 2-RPR serial robot as described above when executing the computer program.

[0022] The present invention also provides a computer-readable storage medium, which stores machine-executable instructions. When the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the parameter optimization design method of the 2-RPR serial robot as described above.

[0023] In general, compared with the prior art, the parameter optimization design method and device of the 2-RPR serial robot provided by the present invention mainly have the following beneficial effects:

[0024] 1. Based on the feasibility judgment results of the workspace and the genetic algorithm, the present invention optimizes the kinematic parameters and branch chain parameters of the 2-RPR robot used for large propeller processing, and considers the obstruction workspace factors caused by the electric cylinder configuration. Under the premise that the workspace meets the requirements, the robot design parameters with better connecting rod length and better branch chain arrangement are obtained, and the branch chain structure is compact.

[0025] 2. Since the robot is used to process large propeller blades, the values ​​of each parameter are relatively large, and the value of fitness f will also be large, which will cause the differences between different individuals to be relatively weakened, and the advantages of roulette selection cannot be brought into play. Therefore, the fitness function is adjusted to give play to the selection advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of a parameter optimization design method of a 2-RPR serial robot provided by the present invention;

[0027] Figure 2 This is a schematic diagram of the structure of a dual electric cylinder high rigidity robot in the present invention;

[0028] Figure 3 is a schematic cross-sectional view of the obstacle space of the robot in the present invention;

[0029] Figure 4 It is a schematic cross-sectional view of the feasibility determination of the working space in the present invention;

[0030] Figure 5 is a schematic diagram of the target working space in the present invention;

[0031] Figure 6 It is a schematic diagram of the target space and the obstacle space in the present invention;

[0032] Figure 7 It is a flow chart of the overlap determination between the target space and the obstacle space in the present invention.

[0033] In all the drawings, the same reference numerals are used to represent the same elements or structures, wherein: 1 - one-axis rotating platform, 2 - two-axis branch chain electric cylinder, 3 - two-axis connecting rod, 4 - three-axis branch chain electric cylinder, 5 - three-axis connecting rod. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0035] See also Figure 1 and Figure 2 The present invention provides a parameter optimization design method for a 2-RPR serial robot, and the parameter optimization design method mainly comprises the following steps:

[0036] Step 1: Multiply the transformation matrices of each link of the 2-RPR serial robot sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the base coordinate system 0, and then substitute the random joint variables into the obtained relative transformation matrix to obtain the robot's workspace.

[0037] Among them, the coordinate system of each link of the robot is established according to the configuration of the 2-RPR serial robot, the link parameters are determined according to the coordinate system of each link, and then the transformation matrices of each link are multiplied sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the polar coordinate system 0, and the random joint variables generated by the random function rand() are substituted into the obtained relative transformation matrix to obtain the coordinate set {P|P(x, y, z)} of the robot end, and then the robot's workspace is obtained.

[0038] In one embodiment, the dual electric cylinder type high rigidity robot for large propeller processing is configured as follows: Figure 2 As shown, the configuration includes a one-axis rotating platform 1, a two-axis chain electric cylinder 2, a two-axis connecting rod 3, a three-axis chain electric cylinder 4 and a robot three-axis connecting rod 5. Different from the traditional six-axis serial robot, the second and third axes of this configuration are driven by RPR type electric cylinder mechanism, which belongs to a high rigidity configuration. The two ends of the two-axis chain electric cylinder 2 are respectively hinged to the one-axis platform and the two-axis connecting rod; the two ends of the three-axis chain electric cylinder 4 are respectively hinged to the two-axis connecting rod and the three-axis connecting rod.

[0039] The acquisition of the robot workspace includes the following sub-steps:

[0040] First, according to the configuration of the 2-RPR serial robot, the coordinate system of each link of the robot is established. The specific steps are as follows:

[0041] (1) Draw the axes of each joint of the robot;

[0042] (2) Draw the common perpendicular line a between two adjacent axes i and i+1 i , find the common perpendicular line a i The intersection point with axis i is defined as the origin O of coordinate system i. i ;

[0043] (3) Provisions i The axis coincides with the axis of joint i;

[0044] (4) Provisions x i Axis and perpendicular line a i Coincident, if z i With z i+1 If they intersect, then x i Yes i and z i+1 The normal to the plane;

[0045] (5) Determine y using the right-hand rule i axis;

[0046] (6) When the first joint variable is zero, coordinate system 0 is required to coincide with coordinate system 1; for the end coordinate system n, the origin and x n The direction can be chosen arbitrarily.

[0047] Next, the connecting rod parameters are determined according to the coordinate system of each connecting rod, specifically:

[0048] a i-1 = from z i-1 to z i Along x i-1 The distance measured;

[0049] α i-1 = from z i-1 to z i Along x i-1 Angle of rotation;

[0050] d i = from x i-1 to x i Along z i The distance measured;

[0051] θ i = from x i-1 to x i Around z i The angle of rotation.

[0052] Next, the transformation matrices of each link are multiplied sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the base coordinate system 0. Among them, the transformation matrix of the link The general expression of is:

[0053]

[0054] By sequentially multiplying the transformation matrices of the links shown in formula (1-1), we can obtain the relative transformation matrix of the end link coordinate system n relative to the base coordinate system 0 (as shown in formula 1-2):

[0055]

[0056] Then, the robot's workspace is solved by the Monte Carlo method, specifically:

[0057] Use the random function rand() to generate random joint variables:

[0058]

[0059] Where q is the joint random variable; q max and q min are the upper and lower limits of the joint angle respectively; i=1,…,n, n is the number of robot joints, and the rotation range of the joint angle of the two-axis and three-axis depends on the stroke of the electric cylinder and the hinge position of the electric cylinder and the connecting rod.

[0060] Then substitute the generated random values ​​of joint angles (joint random variables) into the relative transformation matrix obtained by formula (1-2) to obtain the coordinate set {P|P(x, y, z)} of the robot end, and then draw all the coordinate points in a graph to obtain the point cloud map of the robot's workspace.

[0061] Step 2: Connect the coordinate points of the two joints of the 2-RPR serial robot, the lower hinge point of the electric cylinder, and the upper hinge points of the electric cylinder corresponding to the two extreme positions in sequence to obtain the obstacle space of the robot; wherein, the obstacle space of the robot is the electric cylinder movement area that the robot end can reach in theory but is not allowed to reach in actual use.

[0062] Since the second and third axes of this robot are driven by electric cylinders, unlike traditional robots, there is an area where the electric cylinders move. Although the robot end can reach some areas in theory, it is not allowed to reach them in actual use. This area is called the robot's obstacle space. Therefore, it is necessary to solve the actual size of the obstacle space to ensure that the robot and the workpiece can avoid the obstacle space during processing. The obstacle space caused by the three-axis electric cylinder is not accessible to the robot end, so only the size of the obstacle space caused by the two-axis electric cylinder needs to be considered. The solution method is as follows:

[0063] The horizontal distance from the lower hinge point of the two-axis electric cylinder to the two-axis joint is recorded as c1, and the vertical distance is recorded as e1; the distance from the upper hinge point of the electric cylinder to the two-axis joint is recorded as c2. With the two axes perpendicular to the ground as the zero position, the limit angle of the robot's two axes moving in the negative direction is recorded as θ1, and the limit angle of the robot's two axes moving in the positive direction is recorded as θ2.

[0064] Taking the robot base coordinate as the origin, the coordinates of the two joints are obtained from the DH parameters of the robot (-a1, d1), and the coordinates of the lower hinge point of the electric cylinder are (-a1+c1, d1+e1). As the sleeve rod of the electric cylinder is extended and retracted, the position of the lower hinge point remains unchanged, while the coordinates of the upper hinge point are changing. The coordinates of the upper hinge point of the electric cylinder at the two extreme positions (the longest and shortest extension of the electric cylinder) are (-a1-c2*sinθ1, d1+c2*cosθ1) and (-a1+c2*sinθ2, d1+c2*cosθ2). According to the geometric method, the above four coordinate points are connected in sequence to obtain the obstacle space of the robot, which is composed of a sector area and a triangular area, such as Figure 3 shown.

[0065] Step three, determine whether the robot's workspace envelops the robot's target workspace, and then determine whether the robot's target workspace overlaps with the robot's obstacle space, and determine the coefficient k based on the two judgment results.

[0066] If the robot's workspace envelops the robot's target workspace, and the robot's target workspace does not overlap with the robot's obstacle space, the determination coefficient k is 1; otherwise, the determination coefficient k is 1.5.

[0067] In this embodiment, the workspace feasibility judgment algorithm can compare the robot's workspace with the workspace required for the actual application scenario and output the result of the feasibility judgment. The workspace required for the actual application scenario is referred to as the target workspace. The target workspace is mostly irregular geometric bodies, which are generally abstracted into regular geometric bodies, such as cuboids, cylinders, spheres, etc. For example, a large propeller blade can be replaced by an envelope of a cuboid. The feasibility of the robot's workspace can be judged by whether the characteristic cross-section of the robot's workspace completely covers the characteristic cross-section of the target workspace.

[0068] The robot's workspace is an irregular three-dimensional space, while the target space is a regular geometric body. Therefore, the three-dimensional space envelope problem can be solved by converting it into a two-dimensional characteristic plane envelope problem. For large propeller blades, a rectangular parallelepiped is used as the characteristic geometric body of the blade. When the target workspace is a rectangular parallelepiped, it is only necessary to intercept the three planes Y=0, Z=d1 and Z=d2 in the robot's workspace for judgment, where d1 is the upper surface of the target workspace, d2 is the lower surface of the target workspace, and the three cross sections are as follows: Figure 4 shown.

[0069] Next, we will automatically determine whether the above three sections meet the envelope requirements. The specific steps are as follows:

[0070] For the Y=0 section: Start from the rightmost point of the section, which is the point with the largest x coordinate X max , record the coordinates of the point as (x max ,z0), then when the z coordinate is in the range (z0-H / 2,z0+H / 2) and the x coordinate is less than x max -L / 2, find the point with the largest x coordinate among all the points in the workspace, and record the x coordinate value x0 of the point. At this time, the optimal position of the target workspace in the section can be determined. The coordinates of the four vertices are A(x0,z0+H / 2), B(x0,z0-H / 2), C(x0+L,z0-H / 2), and D(x0+L,z0+H / 2). Figure 5 As shown. Since the values ​​of the horizontal coordinates of points AB are the horizontal coordinate values ​​of the concave points on the left side of the workspace, the line segment formed by AB must be included in the robot workspace, and points CD need to be judged. The judgment method is to find whether the vertical coordinates of the points in the workspace with the horizontal coordinate x0+L / 2 are both smaller than the vertical coordinate of point C and larger than the vertical coordinate of point D. If there are points that meet the requirements, it means that the line segment formed by CD must also be included in the robot workspace, which means that the rectangle formed by the four points ABCD is included in the robot workspace. The judgment conclusion of this section is that it meets the requirements, otherwise it does not meet the requirements.

[0071] For the Z=d1 and Z=d2 sections: the judgment method is the same as the above method steps. If all three sections meet the envelope requirements, it means that the target workspace is contained in the robot workspace, that is, the robot workspace meets the actual use requirements.

[0072] Next, consider the impact of the robot obstacle space: the robot obstacle space is inaccessible at the end in actual processing, so the blade to be processed cannot be placed in the robot obstacle space, that is, the target workspace and the robot obstacle space cannot overlap. It is necessary to determine the coincidence of these two spaces. The specific steps are as follows: Based on the obtained cross-sectional vertex coordinates of the machine obstacle space and the vertex coordinates of the target workspace, according to the shape characteristics of the two, it is only necessary to determine whether point B in the target workspace is in the obstacle space at the Y=0 section, that is, it is only necessary to determine whether point B is directly below the straight line formed by GH, such as Figure 6 The flow chart of the target space and obstacle space coincidence determination is shown in Figure 7 shown.

[0073] There are two output results for the workspace feasibility judgment. If the workspace requirements are met, the output is 1, otherwise the output is 0. k is determined by the workspace feasibility judgment. If the output is 1, then k = 1; if the algorithm output is 0, then k = 1.5.

[0074] Step 4: Use a genetic algorithm to optimize the connecting rod parameters and branch chain parameters of the 2-RPR serial robot; wherein the fitness function of the genetic algorithm is:

[0075] f i '=max{f1,f2,…,f n}-f i +rand()i=1,2,…,n

[0076] In the formula, n is the population size, f i =k*(ω1a2+ω2a3+ω3d4+ω4c1+ω5e1), ω1, ω2, ω3, ω4, ω5 are the weights of the five optimization variables a2, a3, d4, c1, e1 corresponding to population i respectively.

[0077] In this embodiment, according to the aforementioned robot workspace feasibility judgment method, the inclusion relationship between robots of different sizes and target workspaces of different sizes can be quickly determined, and it can be detected whether the target workspace overlaps with the obstacle space, which provides convenience for the subsequent optimization of various robot parameters.

[0078] Genetic algorithm is a search algorithm inspired by the principle of natural evolution. By simulating the reproduction of populations in nature, the phenomena of DNA replication, crossover, and mutation in heredity, the population can evolve to a better state through natural selection of the survival of the fittest. By optimizing the robot parameters through genetic algorithms, the robot parameters with more compact branch chain arrangement and better connecting rod length can be obtained while meeting the requirements of the workspace.

[0079] For the 2-RPR robot, there are three kinematic parameters that affect the robot's workspace, namely a2, a3, and d4. There are five parameters that affect the robot's obstacle space, namely branch chain parameters c1, c2, e1, θ1, and θ2, where θ1 and θ2 are joint rotation angles as invariants, and c2 does not affect the overall size of the robot and is only used to solve the obstacle space.

[0080] Genetic algorithm is used to optimize the above six parameters, including three kinematic parameters a2, a3, d4, and three branch chain parameters c1, c2, e1. The specific steps are as follows:

[0081] First, initialize the genetic algorithm parameters, see Table 1 for details.

[0082] Table 1 Genetic algorithm parameters

[0083]

[0084]

[0085] Then set the fitness function: For the optimization of robot parameters, a major premise is to meet the requirements of the target workspace, and the second optimization goal is to shorten the connecting rod length and adjust the layout of the electric cylinder as much as possible. Therefore, the balance between the two should be fully considered when setting the fitness function. To determine whether the workspace under a set of robot parameters meets the requirements of the target workspace, the above-mentioned workspace judgment method needs to be used. There will be two output results. If the workspace requirements are met, the output is 1, otherwise the output is 0.

[0086] Define the following fitness function:

[0087] f=k*(ω1a2+ω2a3+ω3d4+ω4c1+ω5e1) (4-1)

[0088] In the formula, ω1, ω2, ω3, ω4, and ω5 are the weights of the five optimization quantities, and the optimization strength of different parameters can be adjusted by the user. There is also a variable parameter c2, but it is not an optimization constraint. k is determined by the workspace feasibility judgment algorithm. If the output is 1, k = 1; if the output is 0, k = 1.5; this can well distinguish between good and bad individuals, and the fitness of bad individuals will be higher than that of good individuals.

[0089] Roulette selection is the most common selection method in genetic algorithms. Its basic idea is that the probability of each individual being selected is proportional to its fitness value, that is, the genes of individuals with greater fitness are more likely to be inherited to the next generation, which conforms to the law of natural selection. The specific steps are as follows:

[0090] (1) Calculate the proportion of each individual’s fitness to the total fitness of the population, that is, the probability of being inherited to the next generation.

[0091] (2) Calculate the cumulative probability of each individual.

[0092] (3) Generate a random number and make a selection based on which part of the cumulative probability the random number falls into.

[0093] It is required that the fitness values ​​of different individuals differ significantly, otherwise the difference from random selection is not much. When optimizing the five parameters of the robot, since the robot is used to process large propeller blades, the values ​​of each parameter are relatively large, and the value of fitness f will also be large, which will cause the differences between different individuals to be relatively weakened, and the advantages of roulette selection cannot be brought into play. Therefore, the fitness function needs to be transformed as follows:

[0094] fi '=max{f1,f2,…,f n}-f i +rand() i=1,2,…,n (4-2)

[0095] Where n is the population size. The above formula transforms the fitness values ​​of all individuals, selects the maximum fitness of all individuals and subtracts it from the fitness of each individual. From formula (4-1), we can see that the original fitness of a good individual is small, and after subtraction, a larger fitness value will be obtained, while eliminating the problem of unclear differences caused by excessively high fitness values. A small random amount is added after the difference between the two, in order to retain the worst individual and give it a chance of inheritance.

[0096] Each individual carries six genes, and crossover and mutation operations are also applied to these six genes. By properly adjusting the probabilities of crossover and mutation as well as the degree of gene change during mutation, the genetic algorithm can converge faster.

[0097] The present invention also provides a parameter optimization design system for a 2-RPR serial robot, the system comprising a memory and a processor, the memory storing a computer program, and the processor executing the parameter optimization design method for the 2-RPR serial robot as described above when executing the computer program.

[0098] The present invention also provides a computer-readable storage medium, which stores machine-executable instructions. When the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the parameter optimization design method of the 2-RPR serial robot as described above.

[0099] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A parameter optimization design method for a 2-RPR serial robot, characterized in that: The method comprises the following steps: (1) Connect the coordinate points of the two joints of the 2-RPR serial robot, the lower hinge point of the electric cylinder, and the upper hinge points of the electric cylinder corresponding to the two extreme positions in sequence to obtain the robot's obstacle space; the robot's obstacle space is the electric cylinder movement area that the robot end can theoretically reach but is not allowed to reach in actual use; (2) Determine whether the robot's workspace envelops the robot's target workspace, and then determine whether the robot's target workspace overlaps with the robot's obstacle space, and determine the coefficient k based on the two judgment results; (3) A genetic algorithm is used to optimize the connecting rod parameters and branch chain parameters of the 2-RPR serial robot; wherein the fitness function of the genetic algorithm is: f i '=max{f1,f2,…,f n }-f i +rand()i=1,2,…,j In the formula, j is the population size, f i =k*(ω1a2+ω2a3+ω3d4+ω4c1+ω5e1), ω1, ω2, ω3, ω4, ω5 are the weights of the five optimization variables a2, a3, d4, c1, e1 corresponding to population i respectively.

2. The parameter optimization design method of the 2-RPR serial robot according to claim 1, characterized in that: If the robot's workspace envelops the robot's target workspace, and the robot's target workspace does not overlap with the robot's obstacle space, the determination coefficient k is 1; otherwise, the determination coefficient k is 1.

5.

3. The parameter optimization design method of the 2-RPR serial robot according to claim 1, characterized in that: When judging whether the robot's workspace envelops the robot's target workspace, the three planes Y=0, Z=d1 and Z=d2 in the robot's workspace can be intercepted for judgment.

4. The parameter optimization design method of the 2-RPR serial robot according to claim 3, characterized in that: For the Y=0 section: Start from the rightmost point of the section, which is the point with the largest x coordinate X max , record the coordinates of the point as (x max ,z0), then when the z coordinate is in the range (z0-H / 2,z0+H / 2) and the x coordinate is less than x max -L / 2, find the point with the largest x coordinate among all the points in the workspace within x0+L / 2, and record the x coordinate value x0 of this point. At this time, the optimal position of the target workspace in this section is determined. The coordinates of the four vertices are A(x0,z0+H / 2), B(x0,z0-H / 2), C(x0+L,z0-H / 2), and D(x0+L,z0+H / 2). Since the values ​​of the abscissas of points AB are the abscissas of the concave points on the left contour of the workspace, the line segment formed by AB must be included in the robot workspace, and points CD need to be judged to find out whether the ordinates of the points in the workspace with abscissas x0+L / 2 are both less than the ordinate of point C and greater than the ordinate of point D. If there are points that meet the requirements, it means that the line segment formed by CD must also be included in the robot workspace, which means that the rectangle formed by points ABCD is included in the robot workspace.

5. The parameter optimization design method of the 2-RPR serial robot according to claim 4, characterized in that: According to the shape characteristics of the cross-sectional vertex coordinates of the machine obstacle space and the vertex coordinates of the target workspace, it is only necessary to determine whether point B in the target workspace is within the obstacle space at the Y=0 cross section.

6. The parameter optimization design method of the 2-RPR serial robot according to any one of claims 1 to 5, characterized in that: The transformation matrices of each link of the 2-RPR serial robot are multiplied sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the base coordinate system 0, and then the random joint variables are substituted into the obtained relative transformation matrix to obtain the robot's workspace.

7. The parameter optimization design method of the 2-RPR serial robot according to claim 6, characterized in that: According to the configuration of the 2-RPR serial robot, the coordinate system of each link of the robot is established, and the link parameters are determined according to the coordinate system of each link. Then, the transformation matrices of each link are multiplied sequentially to obtain the relative transformation matrix of the end link coordinate system n relative to the polar coordinate system 0. The random joint variables generated by the random function rand() are substituted into the obtained relative transformation matrix to obtain the coordinate set {P|P(x, y, z)} of the end of the robot, and then the robot's workspace is obtained.

8. The parameter optimization design method of the 2-RPR serial robot according to claim 1, characterized in that: Use the random function rand() to generate random joint variables: Where q is the joint random variable; q max and q min are the upper and lower limits of the joint angle respectively; i=1,…,n, where n is the number of robot joints.

9. A parameter optimization design system for a 2-RPR serial robot, characterized in that: The system includes a memory and a processor, the memory stores a computer program, and the processor executes the parameter optimization design method of the 2-RPR serial robot according to any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores machine-executable instructions. When the machine-executable instructions are called and executed by the processor, the machine-executable instructions prompt the processor to implement the parameter optimization design method for the 2-RPR serial robot according to any one of claims 1 to 8.

Images