Production scheduling integrated optimization method for multiple process routes

Abstract

The invention relates to a multi-process-route-oriented production scheduling integrated optimization method, and relates to the technical field of flexible workshop multi-target scheduling. The method comprises the steps that firstly, a production line process planning and production scheduling integrated optimization model is constructed, the model adapts to a machining sequence flexible scene, and optimal process route determination and procedure machining equipment distribution are synchronously achieved; secondly, double optimization targets are set, namely, the maximum completion time is minimized, and the equipment utilization rate is maximized; thirdly, designing an optimization solution process based on a genetic algorithm framework, generating a chromosome population containing a process route selection machine allocation process sequence, and fusing double targets through an adaptive fitness evaluation mechanism; and finally, developing a visual interaction system in a matched manner, and dynamically outputting an optimal process route scheme and an equipment scheduling plan. The method can effectively solve the problems of long processing task completion time and unbalanced equipment utilization rate in a multi-process route scene, is suitable for a scheduling scene of a flexible job shop, and improves the overall operation efficiency of a production line.

Description

Technical Field

[0001] This invention belongs to the field of flexible workshop scheduling optimization under multiple process routes, and specifically relates to an integrated optimization method for production scheduling under multiple process routes. Background Technology

[0002] Multi-process route production scheduling is a core optimization model for multi-variety, small-batch production in flexible workshops, widely used in discrete manufacturing fields such as aerospace components and precision machining. Its scheduling quality directly determines production efficiency, equipment utilization, and order delivery cycle. Current research on flexible workshop scheduling often focuses on machine allocation and process sequencing under a single process route, or, while considering multiple process routes, separates process route selection from scheduling execution, making it difficult for scheduling schemes to adapt to the complex constraints of multiple workpieces, multiple devices, and multiple routes in actual production. Furthermore, existing methods often neglect the impact of process route selection on subsequent scheduling resource matching, and insufficient consideration is given to equipment utilization and process connection efficiency during optimization. This leads to problems such as equipment overload and excessively long process waiting times in practical applications, failing to meet the dual requirements of flexible workshops for production flexibility and economy. Therefore, there is an urgent need to construct an integrated optimization method that combines process route selection, machine allocation, and process sequencing to improve the overall optimization performance of flexible workshop scheduling. Summary of the Invention

[0003] This invention aims to address the problems of poor adaptability and low efficiency in process route selection, machine allocation, and process sequencing optimization schemes in flexible workshop scheduling. It proposes an integrated optimization method for production scheduling with multiple process routes. Traditional scheduling methods often separate process route selection from subsequent scheduling decisions, or fail to adequately coordinate and optimize equipment load and production cycle, easily leading to resource waste and schedule delays. To address these issues, an integrated optimization framework based on genetic algorithms is proposed, which achieves global optimization by unifying encoding and fusing multi-dimensional decisions.

[0004] This application provides the following technical solutions to achieve the above effects:

[0005] A production scheduling integration optimization method for multiple process routes, the method comprising the following steps:

[0006] Step 1: Construct an integrated optimization model for production line process planning and production scheduling;

[0007] Step 2: Based on the production scheduling integrated optimization model, set dual optimization objectives, namely minimizing the maximum completion time and maximizing equipment utilization, and determine the collaborative optimization requirements of the two objectives; construct a fitness evaluation mechanism to calculate the comprehensive fitness.

[0008] Step 3: Design and optimize the process based on genetic algorithm. By transforming the process route selection, machine allocation, and process sequence into a three-layer structured code, the solution space of the scheduling problem is mapped to the gene sequence of the genetic algorithm. Then, the population is initialized, and the fitness evaluation mechanism is embedded into the selection, crossover, and mutation process of the genetic algorithm to achieve population evolution optimization. Finally, the above specific information is parsed from the gene sequence to form the optimal generation scheduling scheme.

[0009] Furthermore, step 1 specifically includes:

[0010] Step 11: Let the manufacturing task be... ,in This represents the i-th workpiece, where n is the total number of workpieces; each workpiece For multiple feasible process routes, denoted as , Indicates workpiece The j-th process route, where k is the maximum number of process routes per workpiece;

[0011] Step 12: Construct a process route representation method, for each process route. It contains several ordered processes, denoted as , Indicates workpiece The s-th process of route j;

[0012] Step 13: For each process Configure the available equipment set and base processing time; set the processing equipment as... , This represents the p-th device, where m is the total number of devices;

[0013] Step 14: Construct a constraint representation method for production scheduling: process sequence constraint, wherein the process sequence constraint requires that processes for the same workpiece must be performed in accordance with... The sequential execution and equipment constraints in the process are as follows: the equipment constraint is that the same equipment can only execute one process at the same time period; the process route constraint is that each workpiece can only select one process route to execute.

[0014] Furthermore, step 2 specifically includes:

[0015] Step 21: Calculate the maximum completion time of the production task: the maximum completion time of the final process for all workpieces, calculated using the following formula:

[0016]

[0017] in, This represents the completion time of all processes for the i-th workpiece, where N is the total number of workpieces;

[0018] Step 22: Calculate equipment utilization rate:

[0019] Where M represents the total number of devices. This represents the total working time of the p-th device. This represents the maximum completion time of the p-th device;

[0020] Step 23: Normalize the maximum completion time:

[0021] Among them, C max Theoretical maximum completion time:

[0022]

[0023] Where M represents the total number of devices. Indicates workpiece J i The selected process route is r i , This refers to the basic processing time for the corresponding process.

[0024] Step 24: Construct a comprehensive fitness function. By weighted fusion of the above two objectives and adding penalties for completion times and utilization rates exceeding reasonable ranges, calculate the final comprehensive fitness: Where w1 is the completion time weight, w2 is the equipment utilization weight, C is the maximum completion time (normalized value, range [0,1]), U is the equipment utilization, and 1-U is used to transform maximizing utilization into a unified optimization direction that minimizes fitness.

[0025] Furthermore, step 3 specifically involves:

[0026] Step 31: When initializing the population, construct a three-layer coding structure for each chromosome, which includes process route selection, machine allocation, and process sequence.

[0027] Step 32: Generate the initial population. Each chromosome is generated by deep copying the original artifact set and randomly shuffling its order.

[0028] Step 33: Perform selection operations based on elite retention and roulette wheel selection to generate the optimal individual for subsequent genetic operations;

[0029] Step 34: Perform crossover and mutation operations on the processes to gradually optimize the population and eventually converge to the optimal solution.

[0030] Furthermore, the method also includes: Step 4: Develop a visual interactive system, apply the optimization results, realize parameter configuration, process monitoring and result visualization, and output the optimal process route and equipment scheduling plan.

[0031] Furthermore, step 4 specifically involves:

[0032] Step 41: Build a visual interactive system framework, integrating parameter configuration module, process monitoring module and result visualization module, and realize data flow docking with optimization algorithm.

[0033] Step 42: In the parameter configuration module, design a user input interface that supports manual configuration and import of genetic algorithm parameters such as population size, number of iterations, crossover rate, and mutation rate, as well as production data such as workpiece information, equipment parameters, and process route constraints.

[0034] Step 43: The process monitoring module collects key indicators in real time during the genetic algorithm iteration process, including the mean fitness of the population, the optimal fitness value, and the iteration time, and displays the optimization process in real time through dynamic curves, numerical panels, and other forms.

[0035] Step 44: The results visualization module analyzes the optimal solution output by the optimization algorithm, presents the optimal process route as a process-route relationship diagram, and transforms the equipment scheduling plan into a Gantt chart to intuitively display the time arrangement of the process on each equipment and the workpiece processing path.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] 1. This invention provides multiple process routes and a set of alternative equipment. It adopts a three-layer chromosome coding system of route selection, machine allocation, and process sequence, combined with constraint-based automatic route optimization, which improves the flexibility and feasibility of the solution, shortens the completion time, and increases equipment utilization.

[0038] 2. This invention simultaneously calculates and plots the convergence curves of fitness, completion time, and equipment utilization. Combined with elite retention and tournament selection mechanisms, it enables visualized diagnosis and parameter calibration of the solution process, ensuring the acquisition of high-quality scheduling schemes.

[0039] 3. This invention effectively overcomes the shortcomings of traditional genetic algorithms, such as being prone to getting trapped in local optima and having low search efficiency, through multi-task collaboration and adaptive genetic strategies. It is especially suitable for the complex coupling of process route selection and equipment allocation in multi-process route scenarios, and achieves more comprehensive and robust scheduling optimization. Attached Figure Description

[0040] Figure 1 This is a flowchart of the integrated optimization method for production scheduling across multiple process routes proposed in this application.

[0041] Figure 2 Performance index curves for multi-objective optimization algorithms;

[0042] Figure 3This is a comparison diagram of the process routes in the embodiments;

[0043] Figure 4 This is a Gantt chart for the optimal production scheduling scheme. Detailed Implementation

[0044] To achieve the above objectives, the present invention adopts the following technical solution:

[0045] This invention aims to address the challenges of multiple process route selection and equipment constraints in flexible workshop scheduling, proposing an integrated optimization method for production scheduling across multiple process routes. Traditional scheduling methods have two limitations: first, they often fix process routes, ignoring their flexibility and limiting scheduling flexibility; second, they do not adequately consider the balance of equipment utilization, easily leading to some equipment being overloaded while others are idle, ultimately resulting in excessively long completion times and resource waste. To address these issues, this invention designs a multi-objective scheduling strategy based on a genetic algorithm. It generates scheduling schemes through a three-layer coding structure of process route selection, equipment allocation, and process sequence, and introduces a fitness penalty mechanism to strengthen constraints, ultimately achieving the optimal scheduling result with the shortest completion time and the highest equipment utilization.

[0046] This application provides the following technical solutions to achieve the above effects:

[0047] A production scheduling integration optimization method for multiple process routes, the configuration method includes the following steps:

[0048] Step 1: Construct an integrated optimization model for production line process planning and production scheduling;

[0049] Step 2: Set dual optimization objectives, namely minimizing the maximum completion time and maximizing equipment utilization, and clarify the requirements for collaborative optimization of the two objectives; construct an fitness evaluation mechanism and calculate the overall fitness.

[0050] Step 3: Based on the genetic algorithm, design and optimize the process, propose an encoding method that integrates process route selection, machine allocation, and process sequence, and embed the fitness evaluation mechanism into the selection, crossover, and mutation process of the genetic algorithm to achieve population evolution optimization;

[0051] Step 4: Apply the encoding method to the scheduling model, optimize the optimal scheduling scheme based on the genetic algorithm; develop a visual interactive system, apply the optimization results, support parameter configuration, process monitoring and result visualization, and output the optimal process route and equipment scheduling plan.

[0052] This invention is an integrated optimization method for production scheduling based on an adaptive genetic algorithm:

[0053] Genetic algorithms, focusing on flexible workshops, achieve collaborative optimization of process route selection, equipment allocation, and process sequencing through scheduling decisions. This optimization logic can improve equipment utilization while shortening production completion time, providing an efficient and feasible solution for complex production environments. The adaptive genetic strategy employed in the scheduling optimization process dynamically adjusts the genetic algorithm parameters. When population evolution is slow, the algorithm automatically increases the mutation probability to introduce new genes, accelerating the population evolution process and making the optimization effect of genetic algorithms in flexible workshop scheduling more efficient.

[0054] The scheduling scheme is implemented using a three-layer chromosome coding structure to ensure the complete and accurate transmission of all scheduling decision information. Chromosome decoding is the key process for transforming the three-layer coded information into the actual scheduling scheme. The specific steps are as follows:

[0055] S11: Parse the process route selection code, extract the selected route for each workpiece from the process route selection layer of the chromosome, and in the workpiece set In the process, for the i-th workpiece The set of feasible process routes is as follows Each of the routes This corresponds to a separate set of process combinations and processing logic. The process route selection in the chromosome is encoded as a vector. ,in , Indicates workpiece J i Select the r-th vector in R i The route serves as the processing path;

[0056] S12: Parse the process sequence code, for each process route. A process set consists of a series of processes performed in a fixed order, denoted as . ,in Indicates workpiece The j-th process is the s-th step in the j-th route, where t is the total number of steps in this route. After completing the process route selection and equipment allocation, all steps assigned to specific equipment need to be sorted to determine the processing sequence on each equipment. The process sorting in the chromosome is encoded as an ordered sequence containing all steps. ,in This sequence represents a unique identifier for a process. It defines the priority relationship of all processes globally. During actual scheduling, processes on the same device must be processed according to their relative order in sequence S to avoid conflicts between processes and ensure the rationality of the processing logic.

[0057] S13: Parse the equipment allocation code and assign equipment that meets the processing requirements to each process step, simultaneously satisfying both equipment capacity constraints and process requirement constraints. Each process step... Due to processing requirements, the process can only be completed on specific equipment, which constitute the set of optional equipment for this process. ,in Indicate process Let p be the p-th optional device, and t be the number of devices available for this process; all devices involved in production together constitute the device set, denoted as t. , Let p represent the p-th device, and m be the total number of devices; the device allocation encoding in the chromosome is a matrix. Where N is the total number of workpieces. This represents the maximum number of processes in the selected process route for all workpieces;

[0058] S14: Construct a method for representing constraints in production scheduling;

[0059] S14-1: Process Sequence Constraint: Processes along the selected route for the same workpiece must be executed in a fixed order; that is, the next process can only begin after the previous process is completed. For processes... and its subsequent processes , must meet ,in , For process The start time, For process End time, For process In the equipment Processing time;

[0060] S14-2: Equipment Constraint: The same equipment can only process one operation at any given time. (Regarding equipment...) Any two different processes O a and O b If O a Prior to O b Processing requires meeting the following requirements. ;

[0061] S14-3: Process route constraint: Each workpiece can only select one process route for execution, that is, if a route is selected... Then the route includes all the processes. The process on the unselected route will not be included in the scheduling plan.

[0062] Based on the above analytical steps and constraints, a fitness evaluation is required after the scheduling scheme is generated. The fitness function is the core of multi-objective optimization, and its construction process includes three stages: calculation of key indicators, normalization processing, and penalty for out-of-range indicators. All calculation logic is designed around actual production needs.

[0063] The two objectives of fitness assessment are maximum completion time and equipment utilization.

[0064] The calculation of the actual maximum completion time C strictly follows the production constraints for each process. start time Two conditions must be met simultaneously: first, the previous process of the workpiece... The process is complete if two conditions are met: firstly, the assigned equipment is idle; secondly, the process completion time is [not specified]. This is the sum of the start time and the processing time, i.e. During the calculation process, the system continuously monitors for overlapping or conflicting process times. If a conflict is detected, the current process's scheduling is skipped, ensuring that the final completion time is a valid result that meets all constraints. After all processes are scheduled, the time for each machine is first tallied. Final free time Then, take the maximum value of the final idle time of all devices, which is the actual maximum completion time. The calculation formula is as follows:

[0065]

[0066] in, Let N represent the completion time of all processes for the i-th workpiece, where N is the total number of workpieces.

[0067] Equipment utilization rate is calculated based on individual equipment. First, statistics are collected for each piece of equipment. Total working time That is, the processing time of all processes allocated to this equipment. Sum of the results, then calculate the completion time of the equipment itself. The ratio of the two is the utilization rate U of a single device. p Then, the average utilization rate of these devices is calculated using the following formula:

[0068]

[0069] Where M represents the total number of devices. This represents the total working time of the p-th device. This represents the maximum completion time of the p-th device; as the equipment utilization rate index U of the entire scheduling scheme, this index can objectively reflect the utilization efficiency of equipment resources and avoid resource waste due to some equipment being idle.

[0070] To achieve unified optimization across multiple objectives, key performance indicators need to be normalized, and an additional penalty mechanism must be applied. First, the theoretical maximum completion time C is calculated. max The formula is:

[0071]

[0072] Where M represents the total number of devices. Indicates workpiece J i The selected process route is r i , The basic processing time for the corresponding process is defined as Task, which is a set of manufacturing tasks.

[0073] The maximum completion time is then normalized using the following formula:

[0074]

[0075] Standardize it to the range of [0,1] to unify the indicator scale and eliminate the impact of differences in indicator magnitude.

[0076] Penalties are imposed on indicators that exceed reasonable ranges to strengthen optimization constraints. For example, if the maximum completion time exceeds 70% of the theoretical value, the penalized value is [value missing]. The penalty coefficient β is linear with the excess ratio, and the formula is:

[0077]

[0078] The minimum threshold for equipment utilization is 75%, and the maximum threshold is 85%. When equipment utilization... < When this occurs, it indicates that equipment resources are not being fully utilized, and the penalty value formula is:

[0079] When utilization > At times, this may cause equipment to operate under overload, affecting production quality and equipment lifespan. A penalty will also be applied, and the penalty value formula is as follows:

[0080] U represents equipment utilization. 1-U is used to transform maximizing utilization into a unified optimization direction that minimizes fitness, aligning with the completion time objective and facilitating comprehensive fitness calculation.

[0081] The integrated fitness function achieves the fusion of the two objectives through weighted summation, where the completion time weight is set to... =0.6, equipment utilization rate weight set to =0.4, this weight allocation comprehensively considers the manufacturing industry's dual needs for production efficiency and resource utilization, and the fitness function formula is:

[0082] A smaller fitness value F indicates a better scheduling scheme. This function transforms a complex multi-objective optimization problem into a single-objective optimization problem, making it easier for the algorithm to search for the optimal solution.

[0083] The iterative optimization process of the algorithm includes core operations such as population initialization, selection, crossover, and mutation. It also introduces an elite retention strategy and an adaptive mechanism to ensure the efficiency and effectiveness of the optimization. In the population initialization phase, 100 chromosomes are randomly generated according to a three-layer coding rule to form the initial population. Each chromosome is first generated by deep copying the original artifact set, and then its order is randomly shuffled to ensure population diversity. The formula is:

[0084]

[0085] This approach provides diverse starting points for algorithmic search. The selection operation combines elite retention with tournament selection. First, the top 10% of high-quality individuals in the current population are retained. Then, five individuals are randomly selected from the population using tournament selection, and the individual with the best fitness is chosen to enter the next generation. This ensures that high-quality solutions are not lost while maintaining population diversity.

[0086] The crossover operation employs different strategies for each of the three coding layers: single-point crossover for the process route selection code R, uniform crossover for the equipment allocation code M, and ordered crossover for the process sequence code S. This ensures that the offspring chromosomes generated after crossover still meet the scheduling constraints. The crossover probability is set to 0.8 to maintain the population evolution speed while preventing the destruction of high-quality genes. The mutation operation is also designed for the three coding layers, randomly changing the process routes of some workpieces, the equipment allocation for some processes, or the execution order of some processes. The initial mutation probability is 0.2. If the algorithm shows no significant improvement for 5 consecutive generations, the mutation rate automatically doubles, but the maximum does not exceed 0.9. This adaptive adjustment helps the algorithm escape local optima.

[0087] The algorithm has two termination conditions: one is reaching the preset maximum number of iterations. =200 generations, and two, if the algorithm shows no significant improvement for 150 consecutive generations, the algorithm stops iterating and outputs the optimal solution when either condition is met. During the iteration process, the system outputs the current best fitness every 10 generations. The system includes completion time (C) and equipment utilization rate (U), allowing users to monitor and optimize progress in real time. It also records historical data for various indicators, providing support for subsequent visualization analysis.

[0088] The system output includes three parts: optimization metrics, scheduling scheme details, and visualization charts. The optimization metrics mainly include the optimal fitness value. Optimal completion time and optimal equipment utilization These indicators directly reflect the quality of the scheduling plan; the scheduling plan details include the specific process route selection for each workpiece, the equipment allocation for each process, and the start time, end time, and processing time of each process, providing detailed guidance for production execution; the visualization charts present the optimization results in an intuitive graphical way, with the convergence curve including fitness F and completion time. Equipment utilization rate The trend of population diversity with the number of iterations helps users clearly understand the optimization process of the algorithm; the Gantt chart displays the process arrangement on each machine in the form of a horizontal bar chart, with different colors distinguishing different workpieces, the process number is marked on the bar chart, and the chart title clearly indicates the optimal completion time. With optimal equipment utilization This allows users to quickly grasp the time allocation of each process; the process route comparison diagram is designed as a separate sub-diagram for each workpiece, with different colored lines to distinguish between selected and unselected routes, and also marks the available equipment for each process, helping users to intuitively compare the differences between different process routes and the rationality of the selected route.

[0089] The system employs a visual UI design for parameter configuration. All key parameters have preset reasonable default values, while also supporting manual adjustment by users, balancing ease of use and flexibility. Among the production task-related parameters, the number of machines is set to 10 by default, the number of workpieces is configured to 15 by default, each workpiece has 2-3 preset feasible process routes, each route contains 7-9 processes by default, and the basic processing time for each process is randomly set between 10 and 35 minutes. The number of available devices for each process is dynamically allocated based on the total number of machines. If the total number of machines is not less than 2, 2-3 available devices are randomly allocated to each process; if there is only 1 machine, that device is directly allocated. Regarding algorithm parameters, the population size is set to 100 by default, and the maximum number of iterations is... ,

[0090] Cross rate Variation rate Elite retention rate The tournament selection size is fixed at 5. These parameters have been tested and verified multiple times, and can achieve a good balance between search efficiency and solution quality. Users can flexibly adjust them according to the complexity of specific production scenarios.

[0091] The system's UI is built on Tkinter and uses a tabbed design, divided into two core areas: parameter settings and optimization results. The parameter settings area is clearly laid out, with the left side being the basic parameter input area, where users can directly input or adjust various parameters, including the number of machines (m), the number of workpieces (N), and the population size. Maximum number of iterations On the right is the process route setting area, which supports the configuration of process details for multiple workpieces and multiple routes through a scrolling frame. The processing time of each process can be set through a pop-up window using equipment sets, making it easy to operate. The optimization results area contains three independent frames, which are used to display the convergence curves of fitness, completion time and equipment utilization, as well as the Gantt chart of the optimal scheduling scheme and the process route comparison chart. The visualization results are displayed in real time through chart embedding technology.

[0092] The entire system achieves closed-loop operation from parameter input and algorithm optimization to result output through a complete process design. During operation, robustness is ensured through multiple mechanisms, including process conflict detection, index validity checking, chromosome length standardization, and automatic correction of invalid configurations, ensuring stable operation of the algorithm under different input conditions. This method fully leverages the global search advantages of genetic algorithms, combined with scientific encoding methods and fitness function design, effectively solving the multi-objective scheduling optimization problem in flexible workshops. Furthermore, the intuitive visualization interface (displaying convergence curves, Gantt charts, and process route comparison diagrams) enhances the system's usability and interpretability, providing scientific and efficient decision support for manufacturing production scheduling and possessing strong engineering application value.

[0093] Compared with the prior art, the present invention has the following advantages:

[0094] 1. This invention provides multiple process routes and a set of alternative equipment. It adopts a three-layer chromosome coding system of route selection, machine allocation, and process sequence, combined with constraint-based automatic route optimization, which improves the flexibility and feasibility of the solution, shortens the completion time, and increases equipment utilization.

[0095] 2. This invention simultaneously calculates and plots the convergence curves of fitness, completion time, and equipment utilization. Combined with elite retention and tournament selection mechanisms, it enables visualized diagnosis and parameter calibration of the solution process, ensuring the acquisition of high-quality scheduling schemes.

[0096] 3. This invention effectively overcomes the shortcomings of traditional genetic algorithms, such as being prone to getting trapped in local optima and having low search efficiency, through multi-task collaboration and adaptive genetic strategies. It is especially suitable for the complex coupling of process route selection and equipment allocation in multi-process route scenarios, and achieves more comprehensive and robust scheduling optimization.

[0097] Implementation Case:

[0098] The following specific examples illustrate the implementation of the present invention. A machining workshop receives a production task that is broken down into 15 workpieces. The workshop contains 10 available machines, and multiple candidate process routes are provided for each workpiece. The goal is to generate an optimal scheduling scheme with the objectives of minimum and maximum completion time and maximum equipment utilization, including process route selection, equipment allocation, and process sequence arrangement.

[0099] Table 1 shows the equipment attributes, including the equipment number, name, processing steps, and processing time. Table 2 records the processing equipment used for each process of each workpiece. Table 3 shows the occupancy time of each piece of equipment. Table 4 shows the model parameters.

[0100] Table 1 Equipment Attributes and Processable Steps

[0101]

[0102] Table 2. Equipment used for workpiece processing steps

[0103]

[0104] Table 3 Occupancy Time of Each Device

[0105]

[0106] Table 4 Model Parameters

[0107]

[0108] Experimental Analysis:

[0109] To verify the effectiveness of our adaptive genetic strategy-based workshop production scheduling method, we compared it with traditional scheduling methods. Traditional methods do not design specific codes and operational strategies for process route selection, equipment allocation, and process sequence; they only use simple random combinations to generate scheduling schemes. Both methods aim to optimize for minimum and maximum completion times and maximum equipment utilization. Our method generates scheduling schemes through a hierarchical crossover strategy, which involves single-point crossover of process routes, uniform crossover of equipment allocation, and ordered crossover of process sequences, as well as a directional mutation and adaptive mutation rate adjustment mechanism for the three-layer codes. Traditional methods, on the other hand, directly and randomly match process routes, equipment, and process sequences.

[0110] Production scheduling optimization flowchart as follows Figure 1 As shown, the process begins with problem modeling. By defining multiple process routes, setting equipment resource constraints, and clarifying the sequence of processes, a hybrid coding scheme is constructed. Then, the population is initialized, the completion time and equipment utilization rate of each individual are calculated, and fitness is evaluated. Next, a new population is generated through selection, crossover, and mutation operations using a genetic algorithm, and the process is iterated repeatedly until the termination condition is met, finally outputting the optimal solution.

[0111] Figure 2 The study presents the fitness convergence curve, completion time convergence curve, equipment utilization convergence curve, and population diversity change trend.

[0112] Figure 3 The system provides the available process routes for each workpiece, the final selected route, and the available equipment for each process step.

[0113] Optimal production scheduling scheme such as Figure 4 As shown, the total completion time of this scheduling scheme is 372 minutes, and the equipment utilization rate reaches 78.23%.

Claims

1. A production scheduling integrated optimization method for multiple process routes, characterized in that, The method includes the following steps: Step 1: Construct an integrated optimization model for production line process planning and production scheduling; Step 2: Based on the production scheduling integrated optimization model, set dual optimization objectives, namely minimizing the maximum completion time and maximizing equipment utilization, and determine the collaborative optimization requirements of the two objectives; construct a fitness evaluation mechanism to calculate the comprehensive fitness. Step 3: Design and optimize the process based on genetic algorithm. By transforming the process route selection, machine allocation, and process sequence into a three-layer structured code, the solution space of the scheduling problem is mapped to the gene sequence of the genetic algorithm. Then, the population is initialized, and the fitness evaluation mechanism is embedded into the selection, crossover, and mutation process of the genetic algorithm to achieve population evolution optimization. Finally, the above specific information is parsed from the gene sequence to form the optimal generation scheduling scheme.

2. The integrated optimization method for production scheduling across multiple process routes according to claim 1, characterized in that, Step 1 specifically involves: Step 11: Let the manufacturing task be... ,in This represents the i-th workpiece, where n is the total number of workpieces; each workpiece For multiple feasible process routes, denoted as , Indicates workpiece The j-th process route, where k is the maximum number of process routes per workpiece; Step 12: Construct a process route representation method, for each process route. It contains several ordered processes, denoted as , Indicates workpiece The s-th process of route j; Step 13: For each process Configure the available equipment set and base processing time; set the processing equipment as... , This represents the p-th device, where m is the total number of devices; Step 14: Construct a constraint representation method for production scheduling: process sequence constraint, wherein the process sequence constraint requires that processes for the same workpiece must be performed in accordance with... The sequential execution and equipment constraints in the process are as follows: the equipment constraint is that the same equipment can only execute one process at the same time period; the process route constraint is that each workpiece can only select one process route to execute.

3. The integrated optimization method for production scheduling oriented towards multiple process routes according to claim 1, characterized in that, Step 2 specifically involves: Step 21: Calculate the maximum completion time of the production task: the maximum completion time of the final process for all workpieces, calculated using the following formula: ； in, This represents the completion time of all processes for the i-th workpiece, where N is the total number of workpieces; Step 22: Calculate equipment utilization rate: ; Where M represents the total number of devices. This represents the total working time of the p-th device. This represents the maximum completion time of the p-th device; Step 23: Normalize the maximum completion time: ; Among them, C max Theoretical maximum completion time: ； Where M represents the total number of devices. Indicates workpiece J i The selected process route is r i , This refers to the basic processing time for the corresponding process. Step 24: Construct a comprehensive fitness function. By weighted fusion of the above two objectives and adding penalties for completion times and utilization rates exceeding reasonable ranges, calculate the final comprehensive fitness: Where w1 is the completion time weight, w2 is the equipment utilization weight, C is the maximum completion time (normalized value, range [0,1]), U is the equipment utilization, and 1-U is used to transform maximizing utilization into a unified optimization direction that minimizes fitness.

4. The integrated optimization method for production scheduling oriented towards multiple process routes according to claim 1, characterized in that, Step 3 specifically involves: Step 31: When initializing the population, construct a three-layer coding structure for each chromosome, which includes process route selection, machine allocation, and process sequence. Step 32: Generate the initial population. Each chromosome is generated by deep copying the original artifact set and randomly shuffling its order. Step 33: Perform selection operations based on elite retention and roulette wheel selection to generate the optimal individual for subsequent genetic operations; Step 34: Perform crossover and mutation operations on the processes to gradually optimize the population and eventually converge to the optimal solution.

5. The integrated optimization method for production scheduling across multiple process routes according to claim 1, characterized in that, The method further includes: Step 4: Develop a visual interactive system, apply the optimization results, realize parameter configuration, process monitoring and result visualization, and output the optimal process route and equipment scheduling plan.

6. The integrated optimization method for production scheduling oriented towards multiple process routes according to claim 5, characterized in that, Step 4 specifically involves: Step 41: Build a visual interactive system framework, integrating parameter configuration module, process monitoring module and result visualization module, and realize data flow docking with optimization algorithm; Step 42: In the parameter configuration module, design a user input interface to support manual configuration and import of genetic algorithm parameters such as population size, number of iterations, crossover rate, and mutation rate, as well as production data such as workpiece information, equipment parameters, and process route constraints. Step 43: The process monitoring module collects key indicators in real time during the genetic algorithm iteration process, including the mean fitness of the population, the optimal fitness value, and the iteration time, and displays the optimization process in real time through dynamic curves, numerical panels, and other forms. Step 44: The results visualization module analyzes the optimal solution output by the optimization algorithm, presents the optimal process route as a process-route relationship diagram, and transforms the equipment scheduling plan into a Gantt chart to intuitively display the time arrangement of the process on each equipment and the workpiece processing path.

Images