Plate parallel riveting task allocation and path planning collaborative optimization method
By adopting the collaborative optimization method of parallel riveting task allocation and path planning in the riveting machine, the optimal gene encoding and shortest path of rivet holes is automatically allocated by improved adaptive genetic algorithm, the efficiency and quality problems of riveting machine under complex working conditions are solved. , realizes efficient allocation and path planning of riveting tasks, and improves production efficiency and flexibility.
Patent Information
- Application Number
- CN202410963869.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-18
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art is difficult to automate the riveting machine under complex working conditions, resulting in low riveting quality and efficiency. Traditional genetic algorithms are prone to fall into local optimization when solving the shortest riveting path and parallel riveting task allocation, making it difficult to improve the calculation accuracy and convergence speed. .
A collaborative optimization method for parallel riveting task allocation and path planning of plates is proposed. By obtaining the coordinates and station information of rivet holes, calculating the Chebischev distance, establishing a collaborative optimization model, and solving it using improved adaptive genetic algorithms, finally realizing the rivet holes Automatic assignment of optimal gene coding sequences and shortest riveting paths.
The production efficiency and product quality of the riveting machine are improved, the occurrence of leakage and repeated riveting is reduced, the optimal allocation and path planning of riveting tasks are achieved, and the production flexibility and calculation accuracy are improved.
Smart Images

Figure CN118966491A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of assembly process optimization, and in particular to a collaborative optimization method for task allocation and path planning for parallel riveting of plates. Background Art
[0002] Riveting, as one of the three main ways to connect mechanical parts, has an important impact on the overall performance of the equipment. At present, the riveting process is mainly completed by manually operated riveting machines, and fully automated riveting operations are still rare. In order to improve the quality and efficiency of riveting, reduce defects such as missed riveting and repeated riveting, and for complex and diverse riveting conditions, it is of great research significance to develop intelligent riveting machines to replace traditional manual or large riveting equipment to complete riveting tasks.
[0003] In order to meet the needs of mass customized production, it is particularly important to improve the production efficiency and flexible production capacity of riveting machines. Optimizing the riveting path is one of the important means to improve the riveting efficiency. By optimizing the riveting path, the moving distance of the workpiece during the riveting process can be shortened, thereby saving time and energy. The shortest riveting path problem can be analogous to the Traveling Salesman Problem (TSP) that does not return to the starting point, that is, to find the shortest path that passes through all riveting points but does not return to the starting point. The traveling salesman problem is a typical combinatorial optimization problem. The number of possible paths grows exponentially with the number of nodes. When the number of nodes increases, a combinatorial explosion will occur. It is difficult to find the exact optimal solution within a reasonable time, and an effective approximate solution algorithm needs to be adopted.
[0004] In order to further improve the production efficiency of the riveting machine, a multi-station riveting table can be set up to realize parallel riveting processing. In the multi-station parallel riveting process, the reasonable allocation of riveting tasks is very important, which can reduce the production cycle and balance the workload of each station. The parallel riveting task allocation problem belongs to the problem of minimizing the total length of part processing. Since this type of problem is an NP-complete problem, there is no effective analytical algorithm, and it is necessary to iteratively solve the optimal riveting task allocation solution.
[0005] A large number of studies have shown that genetic algorithms have excellent performance in solving discrete combinatorial optimization problems and can be used to solve the shortest riveting path problem and the parallel riveting task allocation problem. However, traditional genetic algorithms still have some shortcomings in solving such problems, such as being prone to falling into local optimality and difficulty in finding the global optimal solution; and slow convergence speed, especially when the problem scale is large. In addition, if the sheet riveting optimization problem of multi-station parallel processing is decomposed into two sub-problems, the shortest riveting path and the parallel riveting task allocation, it may lead to local optimal problems. Therefore, for the sheet riveting optimization problem of multi-station parallel processing, it is necessary to establish a collaborative optimization model and use an improved genetic algorithm to solve it in order to improve the calculation accuracy and convergence speed. Summary of the invention
[0006] The present invention aims to provide a collaborative optimization method for parallel riveting task allocation and path planning of plates, so as to solve the problems of improving the riveting quality and efficiency of intelligent riveting machines under complex working conditions and reducing missed riveting and repeated riveting.
[0007] In order to achieve the above object, the present invention provides the following technical solutions:
[0008] A collaborative optimization method for plate parallel riveting task allocation and path planning includes the following steps:
[0009] S1. Obtain the rivet hole coordinate set and station quantity information;
[0010] S2, eliminating rivet holes that cause fixture interference and obtaining a set of machinable rivet holes;
[0011] S3, calculating the Chebyshev distances between all machinable rivet holes;
[0012] S4. Establish collaborative optimization model;
[0013] S5. Solving collaborative optimization model based on improved adaptive genetic algorithm;
[0014] S6. Perform parallel riveting processing according to the solved optimal gene coding sequence of the rivet hole.
[0015] A further technical feature is that in step S1, the rivet holes in the processing product drawing are automatically identified through CAD analysis, the coordinate system of each rivet hole is unified by coordinate transformation, and then according to the reference position of the actual fixture, the coordinates of the rivet holes in the drawing are transformed into coordinate values that refer to the actual fixture reference.
[0016] Further technical features, in step S3, the shortest riveting path from rivet hole i to j can be expressed by Chebyshev distance as
[0017] d ij =max(x i -x j |,y i -y j |)
[0018] Where i, j are the rivet hole numbers; (x i ,y i ) and (x j ,y j ) are the coordinates of rivet holes i and j respectively; i, j = 1, 2, …, n; n is the total number of rivet holes.
[0019] Further technical features, in step S4, let δ ij is the path planning decision variable, and its expression is as follows
[0020]
[0021] When δ ij = 1, the path from node i to node j can be realized; otherwise, in the considered itinerary, there is no direct connection between nodes, and the shortest riveting path model can be expressed as
[0022] find:x={δ ij}
[0023]
[0024] Where x is the shortest riveted path to be found, f1 represents the riveted path distance, the first and second constraints restrict each node to have only one edge entering and leaving, the third constraint is used to eliminate subpaths, S is all non-empty true subsets in the node set {1,2,...,n}, |S| is the number of nodes contained in the set S;
[0025] A 0-1 integer programming is used to describe the problem of minimizing the total length of parts processing involved in the parallel riveting task allocation of plates. Let f2 represent the total length of riveting processing and define a 0-1 integer variable λ jk Assign decision variables to the riveting task, the expression is as follows
[0026]
[0027] When jk=1 , the rivet hole j is assigned to the workstation k for processing; otherwise, the rivet hole j is not processed at the workstation k. The time required to process the rivet hole j is t j It consists of two parts: equipment moving time and riveting time, which can be expressed as
[0028]
[0029] Where v is the moving speed of the device along the XY axis, d ij δ ij / v represents the variable moving time of the equipment from rivet hole i to j; t0 is a constant value, which represents the fixed riveting time of the equipment except for the movement, including the time of taking pictures, threading pins, and pulling riveting;
[0030] The total length of riveting processing is equal to the processing time required by the workstation with the longest processing time. The optimization model for parallel riveting task allocation of plates can be expressed as
[0031] find:x={λ jk}
[0032]
[0033] Where x is the parallel riveting task allocation scheme with the shortest processing time to be sought, and m is the total number of workstations;
[0034] The collaborative optimization model of plate parallel riveting task allocation and path planning is as follows:
[0035] find:x={δ ij ,λ jk}
[0036]
[0037] Where x is the optimal solution to be sought, including the parallel riveting task allocation scheme with the shortest riveting path and the shortest processing time.
[0038] A further technical feature is that in step S5, the collaborative optimization model is solved based on the improved adaptive genetic algorithm, and the specific steps are as follows:
[0039] S101, establishing an individual coding method for rivet holes and initializing the population;
[0040] S102, establishing a fitness function;
[0041] S103, performing a selection operation according to the individual fitness value;
[0042] S104, calculating the population sample standard deviation, and adaptively adjusting the order of crossover and mutation operations;
[0043] S105, calculating the adaptive crossover and mutation probabilities of the population, and performing crossover and mutation operations;
[0044] S106: Determine whether the stop condition is met.
[0045] Further technical features, in step S101, the individual coding method of the rivet holes is:
[0046] The coding sequence 1, 2, ..., n is used to randomly encode the rivet holes processed by each station, and each rivet hole is processed only once; the breakpoint sequence is used to represent the processing end point of each station in the coding sequence. The sequence is sorted in ascending order, and the sequence length is equal to m-1, where m is the total number of stations.
[0047] Further technical features, in step S102, the fitness function is:
[0048] F=w1f1+(1-w1)f2
[0049] Among them, w1 and 1-w1 are the weight coefficients of objective functions f1 and f2 respectively, and their values reflect the importance of each objective function in the fitness calculation.
[0050] Further technical features, in step S104, the population sample standard deviation is expressed as follows:
[0051]
[0052] In the above formula, r is the population sample standard deviation, F i is the individual fitness, F avg is the average fitness of the population; if the population sample standard deviation r is less than the critical value R, the population is relatively concentrated, and the mutation operation is first performed to change the individual genotype, and then the crossover operation is used to improve the population diversity and jump out of the local optimum; if the population sample standard deviation r is greater than the critical value R, the population is more divergent and has a good population diversity foundation and will not fall into the local optimum, then the crossover operation is performed first, and then the mutation operation is performed.
[0053] A further technical feature is that in step S105, the adaptive crossover probability and mutation probability calculation formula are as follows:
[0054]
[0055] Where P c and P m They represent the adaptive crossover probability and mutation probability, respectively, and the crossover and mutation operations are performed on the population accordingly; and are the maximum and minimum values of the crossover probability set initially, F c is the maximum fitness of both parties of the crossover, and are the maximum and minimum values of the initial mutation probability, F m is the fitness of the mutant individual, F avg is the average fitness of the current population, F max is the maximum fitness of the current population, g is the number of iterations of the current population, and G is the maximum number of iterations of the population.
[0056] The principle and beneficial effects of this technical solution:
[0057] The present invention aims at the optimization problem of plate riveting in multi-station parallel processing, and proposes a collaborative optimization method for riveting task allocation and path planning. The method can automatically allocate the optimal riveting task and the shortest riveting path to each station according to the station quantity information and the rivet hole coordinates automatically obtained from the processed product drawings, thereby achieving a comprehensive improvement in production efficiency, product quality and production flexibility, and providing technical support for the research and development of intelligent riveting machines. BRIEF DESCRIPTION OF THE DRAWINGS
[0058] Figure 1 Schematic diagram of collaborative optimization process for task allocation and path planning for parallel riveting of plates;
[0059] Figure 2 Schematic diagram of the rivet hole gene coding sequence. DETAILED DESCRIPTION
[0060] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:
[0061] like Figure 1 As shown, the present invention provides a collaborative optimization method for plate parallel riveting task allocation and path planning, which specifically includes the following steps:
[0062] S1. Obtain the rivet hole coordinate set and station quantity information.
[0063] First, the production control systems such as MES, ERP and WMS of the manufacturing enterprise are connected through the workshop industrial gateway to obtain the processing product drawings and the number of workstations. Then, the rivet hole elements in the drawings are automatically identified through CAD analysis, and the coordinate systems of the rivet holes in the drawings are unified through coordinate transformation operations such as translation, rotation, and flipping. Finally, according to the reference position of the actual fixture, the rivet hole coordinate set in the drawing is transformed into a coordinate set that refers to the actual fixture reference.
[0064] S2. Eliminate the rivet holes that cause fixture interference and obtain a set of machinable rivet holes.
[0065] From the set of rivet holes in the sheet processing area, remove the rivet holes that interfere with the fixture that grabs the sheet to ensure production safety and processing integrity. The set of machinable rivet holes after removing the rivet holes that cause interference is recorded as V = {1, 2, ..., n}, where n is the total number of rivet holes.
[0066] S3. Calculate the Chebyshev distances between all machinable rivet holes.
[0067] The riveting machine relies on the servo motor of the XY axis to move at a constant speed. The shortest riveting path from rivet hole i to j can be expressed as the Chebyshev distance d ij =max(x i -x j |,y i -y j |), where (x i ,y i ) and (x j ,y j ) are the coordinates of rivet holes i and j, i, j = 1, 2, ..., n.
[0068] By calculating the Chebyshev distances between all machinable rivet holes, a complete graph can be constructed in which each rivet hole is connected to all other rivet holes, and the weight of the edge is the corresponding Chebyshev distance.
[0069] S4. Establish a collaborative optimization model.
[0070] The shortest riveted path problem can be compared to the traveling salesman problem that does not return to the starting point. From the perspective of graph theory, the essence of this problem is to find a Hamilton path with the minimum weight (the starting point and the end point are different) given a weighted completely undirected graph. Let δ ij is the path planning decision variable, and its expression is as follows
[0071]
[0072] When δ ij = 1, the path from node i to node j can be realized; otherwise, in the considered itinerary, there is no direct connection between nodes, and the shortest riveting path model can be expressed as
[0073]
[0074] In formula (2), x is the shortest riveted path to be found, f1 represents the riveted path distance, the first and second constraints restrict each node to have only one edge entering and leaving, the third constraint is used to eliminate subpaths, S is all non-empty true subsets in the node set {1,2,…,n}, and |S| is the number of nodes contained in the set S.
[0075] The problem of minimizing the total length of parts processing involved in the parallel riveting task allocation of plates can be described by a 0-1 integer programming. Let f2 represent the total length of riveting processing and define a 0-1 integer variable λ jk Assign decision variables to the riveting task, the expression is as follows
[0076]
[0077] When jk=1 , the rivet hole j is assigned to the workstation k for processing; otherwise, the rivet hole j is not processed at the workstation k. Since any rivet hole j needs to be processed once at a certain workstation k,
[0078]
[0079] Let t j represents the time required for the equipment to process rivet hole j, t j It consists of two parts: equipment moving time and riveting time, which can be expressed as
[0080]
[0081] Where v is the moving speed of the device along the XY axis, d ij δ ij / v represents the variable moving time of the equipment from rivet hole i to j; t0 is a constant value, which represents the fixed riveting time of the equipment except for movement, including the time for taking pictures, threading pins, and pulling rivets.
[0082] The total length of the riveting process is equal to the processing time required for the longest processing station, that is,
[0083]
[0084] Therefore, the optimization model of plate parallel riveting task allocation can be expressed as
[0085]
[0086] Where x is the parallel riveting task allocation scheme with the shortest processing time to be determined, and m is the total number of workstations.
[0087] From equations (2) and (7), the collaborative optimization model of plate parallel riveting task allocation and path planning can be obtained as follows:
[0088]
[0089] Where x is the optimal solution to be sought, including the parallel riveting task allocation scheme with the shortest riveting path and the shortest processing time. Formula (8) is an NP-complete problem, and there is no effective analytical algorithm (polynomial algorithm), so it needs to be solved by numerical iteration.
[0090] S5. Solve the collaborative optimization model based on the improved adaptive genetic algorithm.
[0091] The improved adaptive genetic algorithm is used to solve the collaborative optimization model of plate parallel riveting task allocation and path planning shown in formula (8). The specific steps are as follows:
[0092] S101, establishing an individual coding method for rivet holes and initializing the population.
[0093] The coding sequence 1, 2, ..., n is used to randomly code the rivet holes processed by each station. Each rivet hole is processed only once. The breakpoint sequence is used to represent the processing end point of each station in the coding sequence. The sequence is sorted in ascending order. The sequence length is equal to m-1, where m is the total number of stations. If m = 3 and n = 8, the coding is as follows Figure 2 The decoding result is that station 1 processes rivet holes 2 and 4, station 2 processes rivet holes 1, 6, and 3, and station 3 processes rivet holes 5, 7, and 8. Figure 2 The rivet hole gene encoding shown only considers the station processing sequence. In the algorithm solution, the starting and ending positions must be added before and after each station processing sequence.
[0094] S102: Establish a fitness function.
[0095] The fitness function corresponds to the objective function in formula (8) and can be expressed as
[0096] F=w1f1+(1-w1)f2 (9)
[0097] Among them, w1 and 1-w1 are the weight coefficients of objective functions f1 and f2 respectively, and their values reflect the importance of each objective function in the fitness calculation.
[0098] S103, performing selection operation according to individual fitness value.
[0099] Individuals in the current population are copied to the new population according to the roulette model with a probability proportional to their fitness.
[0100] S104. Calculate the population sample standard deviation and adaptively adjust the order of crossover and mutation operations.
[0101] Traditional genetic algorithms follow the order of selection, crossover, and mutation. However, when the population is relatively concentrated, crossover is of little significance and the population may still fall into a local optimal solution. Therefore, a sequential adaptive mechanism is introduced to determine the order of crossover and mutation by judging the degree of discreteness of population samples. The population density can be expressed by the sample standard deviation as follows:
[0102]
[0103] In the above formula, r is the population sample standard deviation, F i is the individual fitness, F avg is the average fitness of the population. If the population sample standard deviation r is less than the critical value R, the population is relatively concentrated, and the mutation operation is first performed to change the individual genotype, and then the crossover operation is used to improve the population diversity and jump out of the local optimum; if the population sample standard deviation r is greater than the critical value R, the population is relatively divergent, and there is a good population diversity foundation, and it will not fall into the local optimum, so the crossover operation is performed first, and then the mutation operation is performed.
[0104] S105. Calculate the adaptive crossover and mutation probabilities of the population, and perform crossover and mutation operations.
[0105] The crossover probability and mutation probability in the genetic algorithm have a significant impact on the algorithm's convergence speed and the ability to obtain the optimal solution. However, these two parameters of the traditional genetic algorithm are constants, and it is usually difficult to find their optimal values through repeated experiments. If the crossover probability is too large, it is easy to destroy the structure of high-fitness individuals and reduce the quality of the overall solution; if the crossover probability is too small, it is not easy to destroy the structure of old individuals, which may lead to premature algorithm. If the mutation probability is too small, it is not easy to form new individuals, which reduces the global search ability; if the mutation probability is too large, the randomness is enhanced, which will increase the global search time.
[0106] In order to improve the calculation accuracy and convergence speed of the genetic algorithm, a probability adaptive adjustment function is introduced to make the crossover probability and mutation probability automatically adjusted according to the individual fitness and the number of iterations.
[0107]
[0108] Formulas (11) and (12) are used to calculate the crossover and mutation probabilities of the current population, respectively, and then perform crossover and mutation operations on the population. and are the maximum and minimum values of the crossover probability set initially, F c is the maximum fitness of both parties of the crossover, and are the maximum and minimum values of the initial mutation probability, F m is the fitness of the mutant individual, F avg is the average fitness of the current population, F max is the maximum fitness of the current population, g is the number of iterations of the current population, and G is the maximum number of iterations of the population.
[0109] From equations (11) and (12), we know that when the individual fitness is greater than the current population average fitness, P c and P m It decreases with the increase of individual fitness and population iteration number. Because in the early stage of genetic algorithm, the population needs more diversity to explore the solution space, a larger crossover probability helps to promote new gene combinations, and a larger mutation probability helps to introduce more new genes, increase the diversity of the population, and avoid premature convergence to the local optimal solution. As the iteration proceeds, the algorithm gradually approaches the optimal solution. At this time, the crossover probability and mutation probability are reduced to protect the obtained excellent genes from being destroyed, while allowing subtle search to improve the quality of the solution.
[0110] S106: Determine whether the stop condition is met
[0111] If g<G, the algorithm does not meet the stopping condition, set g=g+1, and jump to S103. Otherwise, the algorithm meets the stopping condition and outputs the optimal gene coding sequence of the rivet hole.
[0112] S6. Perform parallel riveting processing according to the solved optimal gene coding sequence of the rivet hole.
[0113] The improved adaptive genetic algorithm is used to solve the collaborative optimization model shown in formula (8) to obtain the optimal gene coding sequence of rivet holes, which contains the rivet hole processing sequence information of each workstation. The gene order in the processing sequence is the shortest riveting path sorting, based on which efficient parallel riveting processing of the plate can be carried out.
[0114] The present invention aims at the optimization problem of plate riveting in multi-station parallel processing, and proposes a collaborative optimization method for riveting task allocation and path planning. The method can automatically allocate the optimal riveting task and the shortest riveting path to each station according to the station quantity information and the rivet hole coordinates automatically obtained from the processed product drawings, thereby achieving a comprehensive improvement in production efficiency, product quality and production flexibility, and providing technical support for the research and development of intelligent riveting machines.
[0115] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions and / or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, without departing from the technical solution of the present invention, several modifications and improvements can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.
Claims
1. A collaborative optimization method for parallel riveting task allocation and path planning of plates, characterized in that: The following steps are involved: S1. Obtain the rivet hole coordinate set and station quantity information; S2, eliminating rivet holes that cause fixture interference and obtaining a set of machinable rivet holes; S3, calculating the Chebyshev distances between all machinable rivet holes; S4. Establish collaborative optimization model; S5. Solving collaborative optimization model based on improved adaptive genetic algorithm; S6. Perform parallel riveting processing according to the solved optimal gene coding sequence of the rivet hole.
2. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S1, the rivet holes in the product drawing are automatically identified through CAD analysis, the coordinate system of each rivet hole is unified by coordinate transformation, and then the coordinates of the rivet holes in the drawing are transformed into coordinate values referring to the actual fixture reference according to the reference position of the actual fixture.
3. The collaborative optimization method for parallel riveting task allocation and path planning of plates according to claim 1 is characterized by: In step S3, the shortest riveting path from rivet hole i to j can be expressed by Chebyshev distance: d ij =max(x i -x j |,y i -y j |) Where i, j are the rivet hole numbers; (x i ,y i ) and (x j ,y j ) are the coordinates of rivet holes i and j respectively; i, j = 1, 2, …, n; n is the total number of rivet holes.
4. The collaborative optimization method for parallel riveting task allocation and path planning of plates according to claim 1 is characterized by: In step S4, let δ ij is the path planning decision variable, and its expression is as follows When δ ij = 1, the path from node i to node j can be realized; otherwise, in the considered itinerary, there is no direct connection between nodes, and the shortest riveting path model can be expressed as find:x={δ ij } minimize: subject to: Where x is the shortest riveted path to be found, f1 represents the riveted path distance, the first and second constraints restrict each node to have only one edge entering and leaving, the third constraint is used to eliminate subpaths, S is all non-empty true subsets in the node set {1,2,…,n}, |S| is the number of nodes contained in the set S; A 0-1 integer programming is used to describe the problem of minimizing the total length of parts processing involved in the parallel riveting task allocation of plates. Let f2 represent the total length of riveting processing and define a 0-1 integer variable λ jk Assign decision variables to the riveting task, the expression is as follows When jk=1 , the rivet hole j is assigned to the workstation k for processing; otherwise, the rivet hole j is not processed at the workstation k. The time required to process the rivet hole j is t j It consists of two parts: equipment moving time and riveting time, which can be expressed as Where v is the moving speed of the device along the XY axis, d ij δ ij / v represents the variable moving time of the equipment from rivet hole i to j; t0 is a constant value, which represents the fixed riveting time of the equipment except for the movement, including the time of taking pictures, threading pins, and pulling riveting; The total length of riveting processing is equal to the processing time required by the workstation with the longest processing time. The optimization model for parallel riveting task allocation of plates can be expressed as find:x={λ jk } minimize: subject to: Where x is the parallel riveting task allocation scheme with the shortest processing time to be sought, and m is the total number of workstations; The collaborative optimization model of plate parallel riveting task allocation and path planning is as follows: find:x={δ ij ,l jk } minimize: minimize: subject to: Where x is the optimal solution to be sought, including the parallel riveting task allocation scheme with the shortest riveting path and the shortest processing time.
5. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S5, the collaborative optimization model is solved based on the improved adaptive genetic algorithm. The specific steps are as follows: S101, establishing an individual coding method for rivet holes and initializing the population; S102, establishing a fitness function; S103, performing a selection operation according to the individual fitness value; S104, calculating the population sample standard deviation, and adaptively adjusting the order of crossover and mutation operations; S105, calculating the adaptive crossover and mutation probabilities of the population, and performing crossover and mutation operations; S106: Determine whether the stop condition is met.
6. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S101, a coding sequence 1, 2, ..., n is used to randomly encode the rivet holes processed by each station, and each rivet hole is processed only once; a breakpoint sequence is used to represent the processing end point of each station in the coding sequence. The sequence is sorted in ascending order, and the sequence length is equal to m-1, where m is the total number of stations.
7. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S102, the fitness function is: F=w1f1+(1-w1)f2 Among them, w1 and 1-w1 are the weight coefficients of objective functions f1 and f2 respectively, and their values reflect the importance of each objective function in the fitness calculation.
8. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S104, the population sample standard deviation is expressed as follows: In the above formula, r is the population sample standard deviation, F i is the individual fitness, F avg is the average fitness of the population; if the population sample standard deviation r is less than the critical value R, the population is relatively concentrated, and the mutation operation is first performed to change the individual genotype, and then the crossover operation is used to improve the population diversity and jump out of the local optimum; if the population sample standard deviation r is greater than the critical value R, the population is more divergent and has a good population diversity foundation and will not fall into the local optimum, then the crossover operation is performed first, and then the mutation operation is performed.
9. The method for collaborative optimization of plate parallel riveting task allocation and path planning according to claim 1, characterized in that: In step S105, the adaptive crossover probability and mutation probability calculation formula are as follows: Where P c and P m They represent the adaptive crossover probability and mutation probability, respectively, and the crossover and mutation operations are performed on the population accordingly; and are the maximum and minimum values of the crossover probability set initially, F c is the maximum fitness of both parties of the crossover, and are the maximum and minimum values of the initial mutation probability, F m is the fitness of the mutant individual, F avg is the average fitness of the current population, F max is the maximum fitness of the current population, g is the number of iterations of the current population, and G is the maximum number of iterations of the population.
