Shop floor scheduling method based on cuckoo-improved PSO in digital twin environment
By combining the cuckoo algorithm and particle swarm optimization algorithm in a digital twin environment, the problem of insufficient local optimal solution and accuracy in the workshop scheduling is solved, and faster convergence and higher quality scheduling solutions are achieved. , highly adaptable and suitable for complex workshop scheduling issues.
Patent Information
- Application Number
- CN202410446648.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-15
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-04-15
Smart Images

Figure CN118348930B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of workshop scheduling, and in particular relates to a workshop scheduling method based on cuckoo-improved PSO in a digital twin environment. Background Art
[0002] The workshop scheduling problem is a complex problem in the manufacturing industry. It involves how to effectively allocate and utilize resources to ensure the efficiency and smoothness of the production process. In recent years, digital twin technology has become an important tool for optimizing workshop scheduling. Intelligent manufacturing and digital twin technology are closely related. Digital twin technology is an inevitable stage for the in-depth development of intelligent manufacturing. Its core is analysis, reasoning and decision-making, which directly echoes the essential connotation of intelligent manufacturing. Intelligent manufacturing involves a closed-loop process of perception, analysis, reasoning, decision-making and control, while digital twin emphasizes the use of physical models, sensor data, operation history and other information to complete mapping in virtual space through simulation processes to reflect the full life cycle of physical equipment. The development of digital twin technology has gone through three main stages. Initially, it was a period of technology accumulation. Before the 21st century, the emergence of two-dimensional drawing tools and simulation technology laid the foundation for the development of digital twin technology. The second was a period of concept development. Between 2000 and 2015, the basic concept of digital twin was proposed in the aviation and military fields. Finally, it was an explosive application period. Digital twin technology was combined with aerospace, industry, medical and other fields to expand more application scenarios.
[0003] Digital twin technology creates a virtual factory model that can reflect the status of the production line in real time, helping managers make more accurate decisions.
[0004] Among many optimization algorithms, particle swarm optimization (PSO) algorithm is widely used in workshop scheduling problems because of its simplicity and effectiveness. However, it also has some obvious disadvantages in solving workshop scheduling problems.
[0005] First, the PSO algorithm is prone to falling into local optimal solutions, especially when facing complex problems with multiple local extreme points. This is because the algorithm may converge to a non-global optimal solution too early during the search process, thus failing to explore better solutions. This phenomenon is usually caused by the rapid disappearance of the diversity of particles in the algorithm, causing the entire particle swarm to gather around the local optimal solution too early.
[0006] Secondly, the PSO algorithm is not accurate enough. Since the algorithm lacks the cooperation of fine search methods, it often cannot obtain very accurate results. In each iteration, the PSO algorithm only uses the information of the group optimum and the individual optimum, and does not make full use of other information obtained during the calculation process.
[0007] In addition, although the PSO algorithm provides the possibility of global search, it cannot guarantee convergence to the global optimal point. This is particularly important in the workshop scheduling problem, because finding the global optimal solution is crucial to optimizing production efficiency and reducing costs.
[0008] In addition, the PSO algorithm does not have a strict theoretical basis at present. It is mainly designed by simplifying the simulation of a certain group search phenomenon, but it does not explain in principle why this algorithm is effective and its scope of application. This limits the improvement and application of the algorithm to a certain extent. Summary of the invention
[0009] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a workshop scheduling method based on cuckoo improved PSO in a digital twin environment with strong global search capability, fast convergence speed and strong adaptability.
[0010] The technical solution adopted to solve the above technical problems is:
[0011] A workshop scheduling method based on cuckoo improved PSO in a digital twin environment includes the following steps:
[0012] Step 1. Particle swarm initialization
[0013] Define the objective function of the workshop scheduling problem, randomly generate the initial particle swarm according to the workshop scheduling problem, each particle represents a potential scheduling solution, and assign a random speed to each particle to indicate the changing trend of the scheduling solution;
[0014] Step 2. Design a coding scheme to map the scheduling scheme into the position vector of the particle;
[0015] Step 3. Construct a fitness function based on the objective function to evaluate the pros and cons of the scheduling scheme represented by each particle;
[0016] Step 4. Determination of individual optimum and global optimum
[0017] Calculate the fitness of each particle and record the individual optimal solution of each particle, and find the particle with the best fitness among all particles as the global optimal solution;
[0018] Step 5. Parameter setting and adaptive adjustment
[0019] The parameters include inertia factor w, acceleration constants c1 and c2, maximum speed V max and speed minimum V min , number of iterations T, particle swarm size N;
[0020] Adaptive adjustment includes dynamic adjustment of inertia factor, fitness feedback adjustment, iteration number adjustment, parameter range adjustment, and performance-based adjustment during the search process;
[0021] Step 6. Particle update based on cuckoo strategy
[0022] The particle velocity is updated as follows:
[0023]
[0024] In the formula, For particle x i The new velocity in the dth dimension, For particle x i The current velocity in the dth dimension, For particle x i The current position in the dth dimension, For particle x i The optimal position of an individual in the dth dimension, R1 and R2 are random numbers between [0,1], used to introduce randomness, and F is a disturbance factor, used to control the influence intensity of the cuckoo strategy. are the positions of other randomly selected particles, representing the random exploration behavior in the cuckoo strategy;
[0025] Update the particle position according to the following formula:
[0026]
[0027] In the formula, For particle x i The new position in the dth dimension;
[0028] Step 7. Crossover and mutation operations
[0029] Crossover operation: select two or more particles from the current particle group as crossover objects, randomly select one or more crossover points in the particle position vector, and exchange the information of the crossover object at the crossover point according to the selected crossover method. The result of the crossover is one or more new position vectors, which represent the new scheduling scheme.
[0030] Mutation operation: select one or more particles from the current particle swarm as mutation objects, select a mutation method, and modify the position information of the mutation object;
[0031] Step 8. Iteration process and convergence judgment
[0032] Repeat steps 3 to 7 to iteratively update the particle swarm. After each iteration, determine whether it has converged, that is, whether a satisfactory solution has been found or the maximum number of iterations has been reached;
[0033] Step 9. Output the optimal solution
[0034] At the end of the iteration, the scheduling scheme corresponding to the particle with the best fitness value is output as the optimal solution, and the optimal solution is decoded and sorted to meet the needs of actual workshop scheduling.
[0035] Preferably, the encoding scheme in step 2 is one of sequential encoding, permutation encoding, binary encoding, real number encoding, mixed encoding, and random key-based encoding.
[0036] Preferably, the method for constructing the fitness function in step 3 is: based on the objective function of the workshop scheduling problem, determine the constraints of the workshop scheduling problem, integrate the constraints into the objective function, define the fitness function to evaluate the objective function, and score each scheduling scheme. If the scheduling scheme violates the constraints, reduce its fitness score by introducing a penalty term. The penalty term should be large enough to ensure that scheduling schemes that violate the constraints are effectively avoided during the search process, and the fitness score is normalized to guide the search behavior of the particles.
[0037] Preferably, in step 5, the value of the inertia factor w is 0.4-0.9, and the values of the acceleration constants c1 and c2 are both 2.
[0038] Preferably, the crossover method in step 7 is single-point crossover, multi-point crossover or uniform crossover.
[0039] Preferably, the mutation method in step 7 is random reset or exchange mutation or perturbation mutation.
[0040] Preferably, the method for dynamically adjusting the inertia factor in step 5 is:
[0041]
[0042] In the formula, w min and w max are the preset minimum and maximum inertia factors, w min =0.4, w max =0.9, is the fitness of particle xi at the tth iteration, is the average fitness of all particles at the tth iteration, is the minimum fitness of all particles at the tth iteration.
[0043] A computer-readable storage medium stores a computer program / instruction, which, when executed by a processor, implements the steps of a workshop scheduling method based on a cuckoo-improved particle swarm optimization algorithm in a digital twin environment.
[0044] A computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps of a workshop scheduling method based on a cuckoo-improved particle swarm optimization algorithm in a digital twin environment.
[0045] The beneficial effects of the present invention are as follows:
[0046] The present invention combines the advantages of the cuckoo search algorithm and the particle swarm optimization algorithm to solve complex optimization problems, especially workshop scheduling problems, and has the following advantages:
[0047] 1) Enhanced global search capabilities
[0048] The introduction of the cuckoo algorithm increases the breadth of the search space, so that the present invention can explore the solution space more comprehensively and improve the possibility of finding the global optimal solution.
[0049] 2) Improve convergence speed
[0050] The particle swarm optimization algorithm is characterized by rapid convergence, and the addition of the cuckoo algorithm further accelerates the convergence process, so that the present invention can find a high-quality solution in a shorter time.
[0051] 3) Avoid premature convergence
[0052] The random exploration characteristic of the cuckoo algorithm helps the algorithm to jump out of the local optimal solution, reducing the risk of premature convergence in the case of insufficient search.
[0053] 4) Improve population diversity
[0054] The cuckoo algorithm increases the diversity of the population by randomly selecting and replacing particles, which helps to prevent the present invention from falling into a single search path.
[0055] 5) Improve optimization effect
[0056] The present invention combines the advantages of the two algorithms, and can improve the quality of solutions and search efficiency through the global search capability of the cuckoo algorithm while maintaining the characteristics of the PSO algorithm being easy to implement and fast to converge.
[0057] 6) Strong adaptability
[0058] The present invention can adapt to different optimization problems and changing environmental conditions, and has good adaptability to dynamic changes and uncertainties in workshop scheduling.
[0059] 7) Flexibility and scalability
[0060] The structure of the present invention allows adjustment and expansion according to the characteristics of specific problems, for example, by introducing specific encoding and decoding strategies to adapt to the special requirements of shop scheduling.
[0061] 8) Reduce computing resource consumption
[0062] Through effective search strategies and adaptive parameter adjustment, the present invention can find excellent solutions within a relatively small number of iterations, thereby reducing the consumption of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Figure 1 It is a schematic diagram of the process of the present invention.
[0064] Figure 2 It is a comparison chart of the results of the present invention and the existing algorithm in the Penalized function test.
[0065] Figure 3 This is a performance comparison chart of the present invention and other algorithms in the VFR test set test.
[0066] Figure 4 It is the Gantt chart of the workshop scheduling in the 30*15 example of the present invention. DETAILED DESCRIPTION
[0067] The present invention will be further described in detail below in conjunction with the accompanying drawings and examples, but the present invention is not limited to the following embodiments.
[0068] exist Figure 1 In the digital twin environment of this embodiment, the workshop scheduling method based on cuckoo improved PSO includes the following steps:
[0069] Step 1. Particle swarm initialization
[0070] Step 1.1. Determine the objective function
[0071] Clarify the goal of the workshop scheduling problem, such as minimizing completion time, cost or maximizing machine utilization; establish an objective function based on the goal. The goal of this embodiment is to minimize the maximum completion time, and the objective function is the maximum value of the completion time of all jobs; determine the constraints of the workshop scheduling problem, such as each machine can only process one job at a time, each job must be processed in a given process sequence, etc., incorporate the constraints into the objective function, and use penalty terms to handle violations of the constraints.
[0072] Mathematically described as: n workpieces to be processed need to be processed by m processing machines, and the processing process needs to meet the following constraints: the workpiece to be processed is processed only once on each machine; each workpiece is processed in the same order on each machine; all workpieces on each machine are processed in the same order; each machine can only process one workpiece at a time. The goal is to find the workpiece processing order corresponding to one or more optimal target values. Consider the permutation flow shop scheduling problem with the goal of minimizing the maximum completion time. Denote p i,jis the processing time of the jth step of workpiece i on the machine, assuming that the processing preparation time of each workpiece is zero or is included in the processing time p i,j Inner, C(j i , k) is the processing completion time of the jth step of workpiece i on machine k. Without loss of generality, assuming that each workpiece is processed in the order of machine 1 to machine m, the mathematical model of the processing completion time of each workpiece on each machine is as follows:
[0073]
[0074]
[0075]
[0076]
[0077] C max (π)=C(j n ,m) (5)
[0078] Formula (5) is the maximum completion time, and formula (6) is the processing order of the workpiece when the completion time is the minimum. In the formula, π is the scheduling plan, that is, the processing order of a workpiece, π * is the optimal scheduling scheme, that is, the optimal processing sequence of workpieces, and Π is the set of all scheduling schemes, that is, all possible arrangements of the processing sequence of all workpieces.
[0079] Step 1.2. Generate initial particle swarm
[0080] According to the objective function of the workshop scheduling problem, an initial particle swarm is randomly generated, each particle represents a potential scheduling solution, and the dimension of each particle is determined, that is, the number of elements contained in the scheduling solution represented by each particle. For example, if there are m machines and n workpieces to be processed in the workshop, the dimension of the particle may be a sequence containing n elements, each element represents the processing order of a workpiece to be processed on m machines. A random speed is assigned to each particle to indicate the changing trend of the scheduling solution.
[0081] Step 2. Design a coding scheme to map the scheduling scheme into the position vector of the particle;
[0082] Among them, the coding scheme is one of sequential coding, permutation coding, binary coding, real number coding, mixed coding, and random key-based coding. The coding scheme directly affects the representation of the solution and the structure of the search space. The appropriate coding scheme is selected according to the specific characteristics of the workshop scheduling problem, such as the number of jobs, the number of machines, the process sequence, resource constraints, etc. The coding scheme should be able to ensure that each element of the particle position vector can effectively reflect the key features of the scheduling plan and can be easily converted back to the actual scheduling plan.
[0083] Among them, this embodiment adopts the SPV (Smallest Positon Value) rule encoding based on random keys to generalize the basic PSO algorithm to the discrete scheduling problem. As shown in Table 1, the particle position vector Xi = [0.51, -0.72, -0.87, 0.47, -0.98] is mapped to the scheduling scheme π = [5, 3, 2, 4, 1] using the SPV rule.
[0084] Table 1 SPV mapping rules
[0085] Xi 0.51 -0.72 -0.87 0.47 -0.98 Π 5 3 2 4 1
[0086] When this paper adopts the encoding method based on the SPV rule, there may be one or more identical values in the individual particle position vectors. In this case, as the position increases, the values at these positions can be added with a sufficiently small positive number to make the values of each particle position component different. In addition, by using double-precision real numbers to define the particle position components, it is almost impossible for multiple identical values to exist in the particle position vector.
[0087] Step 3. Construct a fitness function based on the objective function to evaluate the pros and cons of the scheduling scheme represented by each particle;
[0088] The method of constructing the fitness function is: based on the objective function of the workshop scheduling problem, determine the constraints of the workshop scheduling problem, integrate the constraints into the objective function, define the fitness function to evaluate the objective function, and score each scheduling plan. If the scheduling plan violates the constraints, reduce its fitness score by introducing a penalty term. The penalty term should be large enough to ensure that scheduling plans that violate the constraints are effectively avoided during the search process. The fitness score is normalized to better guide the search behavior of particles.
[0089] The fitness function f of this embodiment is:
[0090]
[0091] In the formula, x i is a particle, ε is a constant used to avoid division by zero, P(x i ) is particle xi The degree of constraint violation, β is the penalty coefficient, which is used to adjust the fitness score of the constraint violation solution.
[0092] Step 4. Determination of individual optimum and global optimum
[0093] Calculate the fitness of each particle and record the individual optimal solution of each particle, and find the particle with the best fitness among all particles as the global optimal solution;
[0094] Step 5. Parameter setting and adaptive adjustment
[0095] The parameters include inertia factor w, acceleration constants c1 and c2, maximum speed V max and the minimum speed V min , number of iterations T, particle swarm size N; the value of the inertia factor w is 0.4-0.9, which determines the influence of the particle to maintain the previous speed. The initial value is set to 0.7 or 0.8 to balance the global and local search capabilities at the beginning of the search; the values of the acceleration constants c1 and c2 are both 2, which determines the tendency of the particle to approach the individual optimal solution and the global optimal solution; the maximum speed V max and the minimum speed V mi n is determined according to the characteristics of the problem and the size of the search space; the number of iterations T is set according to the complexity of the problem and the expected convergence speed to avoid infinite search; the particle swarm size N is determined according to the scale of the problem and computing resources.
[0096] Adaptive adjustment includes dynamic adjustment of inertia factor, fitness feedback adjustment, iteration number adjustment, parameter range adjustment, and performance-based adjustment during the search process; among them, dynamic adjustment of inertia factor, for example, if it is found that the particle swarm converges to the local optimal solution too early, the inertia factor can be reduced to enhance the local search capability.
[0097] The method for dynamically adjusting the inertia factor is:
[0098]
[0099] In the formula, w min and w max are the preset minimum and maximum inertia factors, w min =0.4, w max =0.9, is the fitness of particle xi at the tth iteration, is the average fitness of all particles at the tth iteration, is the minimum fitness of all particles at the tth iteration.
[0100] Fitness feedback adjustment is to adjust parameters according to the overall fitness feedback of the particle swarm. If the quality of the optimal solution is found to be stagnant, the search capability can be enhanced by adjusting the acceleration constant. The number of iterations adjustment means that if a satisfactory solution is found before the maximum number of iterations is reached, the search can be terminated early. Conversely, if the search progresses slowly, you can consider increasing the number of iterations. Parameter range adjustment is to dynamically adjust the range of speed limits during the search process to adapt to changes in the search space. Performance-based adjustment uses the performance indicators of the algorithm, such as the quality of the solution and the convergence speed, to guide the adjustment of parameters. For example, if the quality of the solution improves slowly, you can try to increase the acceleration constant.
[0101] Step 6. Particle update based on cuckoo strategy
[0102] The particle velocity is updated as follows:
[0103]
[0104] In the formula, For particle x i The new velocity in the dth dimension, For particle x i The current velocity in the dth dimension, For particle x i The current position in the dth dimension, For particle x i The optimal position of an individual in the dth dimension, R1 and R2 are random numbers between [0,1], used to introduce randomness, and F is a disturbance factor, used to control the influence intensity of the cuckoo strategy. are the positions of other randomly selected particles, representing the random exploration behavior in the cuckoo strategy;
[0105] Update the particle position according to the following formula:
[0106]
[0107] In the formula, For particle x i The new position in the dth dimension;
[0108] Step 7. Crossover and mutation operations
[0109] Crossover operation: select two or more particles from the current particle group as crossover objects, randomly select one or more crossover points in the particle position vector, and the crossover mode is single-point crossover, multi-point crossover or uniform crossover. According to the selected crossover mode, exchange the information of the crossover object at the crossover point. The result of the crossover is one or more new position vectors, which represent the new scheduling scheme.
[0110] Mutation operation: select one or more particles from the current particle swarm as mutation objects, select a mutation method, such as random reset, exchange mutation, or perturbation mutation, and modify the position information of the mutation object; for example, random reset mutation assigns a new random position to the particle, exchange mutation swaps the positions of two elements in the position vector; perturbation mutation adds a small random perturbation near the current position.
[0111] Step 8. Iteration process and convergence judgment
[0112] Repeat steps 3 to 7 to iteratively update the particle swarm. After each iteration, determine whether it has converged, that is, whether a satisfactory solution has been found or the maximum number of iterations has been reached;
[0113] Step 9. Output the optimal solution
[0114] At the end of the iteration, the scheduling scheme corresponding to the particle with the best fitness value is output as the optimal solution, and the optimal solution is decoded and sorted to meet the needs of actual workshop scheduling.
[0115] The above method can effectively solve complex workshop scheduling problems, provide high-quality scheduling solutions, optimize production processes, and improve resource utilization and production efficiency.
[0116] A computer-readable storage medium stores a computer program / instruction, which, when executed by a processor, implements the steps of a workshop scheduling method based on cuckoo-improved PSO in a digital twin environment.
[0117] A computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps of a workshop scheduling method based on cuckoo improved PSO in a digital twin environment.
[0118] experiment
[0119] Many complex functions were found for testing the present invention, and some test results are shown below:
[0120] 1. Penalized function test
[0121] One or more Penalized functions with multiple local optimal solutions are selected as test objects. Existing algorithms include mean particle swarm optimization algorithm (MeanPSO), time-varying coefficient adaptive particle swarm optimization algorithm (TCSPSO algorithm), linear decreasing weight particle swarm optimization algorithm (LDWPSO), self-learning particle swarm optimization algorithm (SLPSO), parallel hybrid particle swarm optimization algorithm (PHSPSO). The present invention is compared with different existing algorithms, such as Figure 2, 100-dimensional space iteration results show that the present invention is far superior to other algorithms in terms of both the number of iterations and the optimal result. By combining the global search capability of the cuckoo algorithm and the fast convergence characteristics of the particle swarm algorithm, the present invention can effectively avoid falling into the local optimal solution, thereby achieving better performance on complex optimization problems. The superiority and reliability of the present invention in solving challenging optimization problems are fully verified, providing confidence in its application to actual workshop scheduling problems.
[0122] 2. VFR test set test
[0123] In the VFR test set, 30 different test sets were tested for replacement workshop scheduling. The superiority of the existing algorithms and the present invention was compared according to several standards. Among them, the index BRE (Best Relative Error) represents the percentage error of the optimal solution of the problem obtained by the algorithm relative to the known optimal solution of the problem, the index ARE (Average Relative Error) represents the percentage error of the average value of the solution of the problem obtained by the algorithm relative to the known optimal solution of the problem, and the index WRE (Worst Relative Error) represents the percentage error of the worst solution of the problem obtained by the algorithm relative to the known optimal solution of the problem. The comparison data is as follows Figure 3 As shown, the three indicators of the present invention are obviously better than those of other algorithms.
[0124] For the 30*15 example case in the VFR test set, the present invention found a scheduling sequence close to the official optimal solution under the conditions of a population size of 30 and 50 iterations. The maximum completion time of the present invention is 13045. The optimal time is 2510. The scheduling sequence after optimization by the present invention is: [21,29,11,2,12,24,4,22,28,30,20,18,19,5,16,23,1,9,10,26,14,13,27,7,3,15,6,25,17,8]. The workshop is arranged for production and processing in the above order, and the time taken is 2513, which is very close to the official optimal solution, achieving high stability, short time consumption, and reliable optimal solution data. The workshop processing flow is designed according to the iterative results, such as Figure 4 As shown in the figure, the optimized workshop scheduling process is demonstrated, which clearly shows the effectiveness of the algorithm in practical applications.
[0125] The present invention supports online dynamic adjustment and monitoring, can adapt to interference and failures in workshop operation, and quickly find the optimal scheduling solution.
[0126] Through these experimental processes and results, it can be seen that the present invention has significant advantages in solving the workshop scheduling problem, can effectively find high-quality solutions, and has strong adaptability and stability.
Claims
1. A workshop scheduling method based on cuckoo improved PSO in a digital twin environment, characterized in that: The following steps are involved: Step 1. Particle swarm initialization Define the objective function of the workshop scheduling problem, randomly generate the initial particle swarm according to the workshop scheduling problem, each particle represents a potential scheduling solution, and assign a random speed to each particle to indicate the changing trend of the scheduling solution; Step 2. Design a coding scheme to map the scheduling scheme into the position vector of the particle; Step 3. Construct a fitness function based on the objective function to evaluate the pros and cons of the scheduling scheme represented by each particle; Step 4. Determination of individual optimum and global optimum Calculate the fitness of each particle and record the individual optimal solution of each particle, and find the particle with the best fitness among all particles as the global optimal solution; Step 5. Parameter setting and adaptive adjustment The parameters include inertia factor w, acceleration constants c1 and c2, maximum speed V max and speed minimum V min , number of iterations T, particle swarm size N; Adaptive adjustment includes dynamic adjustment of inertia factor, fitness feedback adjustment, iteration number adjustment, parameter range adjustment, and performance-based adjustment during the search process; Step 6. Particle update based on cuckoo strategy The particle velocity is updated as follows: In the formula, For particle x i The new velocity in the dth dimension, For particle x i The current velocity in the dth dimension, For particle x i The current position in the dth dimension, For particle x i The optimal position of an individual in the dth dimension, R1 and R2 are random numbers between [0,1], used to introduce randomness, and F is a disturbance factor, used to control the influence intensity of the cuckoo strategy. are the positions of other randomly selected particles, representing the random exploration behavior in the cuckoo strategy; Update the particle position according to the following formula: In the formula, For particle x i The new position in the dth dimension; Step 7. Crossover and mutation operations Crossover operation: select two or more particles from the current particle group as crossover objects, randomly select one or more crossover points in the particle position vector, and exchange the information of the crossover object at the crossover point according to the selected crossover method. The result of the crossover is one or more new position vectors, which represent the new scheduling scheme. Mutation operation: select one or more particles from the current particle swarm as mutation objects, select a mutation method, and modify the position information of the mutation object; Step 8. Iteration process and convergence judgment Repeat steps 3 to 7 to iteratively update the particle swarm. After each iteration, determine whether it has converged, that is, whether a satisfactory solution has been found or the maximum number of iterations has been reached; Step 9. Output the optimal solution At the end of the iteration, the scheduling scheme corresponding to the particle with the best fitness value is output as the optimal solution, and the optimal solution is decoded and sorted to meet the needs of actual workshop scheduling.
2. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The encoding scheme in step 2 is one of sequential encoding, permutation encoding, binary encoding, real number encoding, mixed encoding, and random key-based encoding.
3. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The method for constructing the fitness function in step 3 is: based on the objective function of the workshop scheduling problem, determine the constraints of the workshop scheduling problem, integrate the constraints into the objective function, define the fitness function to evaluate the objective function, and score each scheduling plan. If the scheduling plan violates the constraints, its fitness score is reduced by introducing a penalty term. The penalty term ensures that the scheduling plan that violates the constraints is avoided during the search process. The fitness score is normalized to guide the search behavior of the particle.
4. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The value of the inertia factor w in step 5 is 0.4 to 0.
9.
5. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: In step 5, the values of the acceleration constants c1 and c2 are both 2.
6. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The crossover method in step 7 is single-point crossover, multi-point crossover or uniform crossover.
7. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The mutation method in step 7 is random reset or exchange mutation or disturbance mutation.
8. The workshop scheduling method based on cuckoo improved PSO in the digital twin environment according to claim 1 is characterized in that: The method for dynamically adjusting the inertia factor in step 5 is: In the formula, w min and w max are the preset minimum and maximum inertia factors, w min =0.4, w max =0.9, is the particle x at the tth iteration i The fitness of is the average fitness of all particles at the tth iteration, is the minimum fitness of all particles at the tth iteration.
9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.
10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.
