Railway vehicle product multi-variety intelligent optimization automatic scheduling method based on SMART system

Abstract

The invention relates to a rail vehicle product multi-variety intelligent optimization automatic scheduling method based on an SMART system, relates to the technical field of rail vehicle manufacturing, and solves the technical problems of low rail vehicle multi-variety production scheduling efficiency, unreasonable resource allocation, poor accuracy, weak strain capacity and insufficient visualization and traceability in the prior art. The automatic scheduling method comprises the following steps: modeling scheduling basic data based on SMART; constructing a multi-constraint intelligent scheduling model; scheduling optimization solution of the improved genetic algorithm; a dynamic adjustment and real-time monitoring mechanism; designing a visual human-computer interaction interface; and issuing a work order and executing a closed loop. According to the automatic scheduling method, the scheduling efficiency can be improved, the resource configuration can be optimized, the scheduling accuracy can be guaranteed, the dynamic strain capacity can be enhanced, and visual management and control can be realized.

Description

Technical Field

[0001] This invention relates to the field of rail vehicle manufacturing technology, and in particular to a multi-variety intelligent optimization automatic scheduling method for rail vehicle products based on the SMART system. Background Technology

[0002] The production of rail vehicles involves the production of a wide variety of products. Different models and configurations of rail vehicles differ in production processes, production times, and resource requirements. Traditional scheduling methods rely heavily on manual experience, which is inefficient, error-prone, and unable to meet the demands of multi-variety, small-batch production. Intensifying competition in the rail vehicle market and increasingly stringent customer requirements for product delivery cycles and quality necessitate that manufacturers can more efficiently plan production, rationally allocate resources, improve production efficiency, and reduce production costs. Existing systems have limitations in scheduling functionality, lacking intelligent optimization algorithms and effective support for multi-variety product production. They cannot fully utilize production data for in-depth analysis and decision-making, making it difficult to achieve efficient scheduling for multi-variety rail vehicle production.

[0003] In general, existing methods for scheduling rail vehicle products have the following drawbacks: (1) Low scheduling efficiency: Manual scheduling takes 2-3 days to complete the planning of a single batch of multi-variety orders. For multi-process components such as bogies, the process connection is prone to confusion, which leads to extended production waiting time. (2) Imbalance in resource allocation: Due to insufficient coordination of production line capacity, site constraints and rest time allocation, some production lines often operate at overload while others are idle. According to statistics, the utilization rate of special equipment under the traditional scheduling model is only 60%-70%. (3) Poor scheduling accuracy: Manual experience judgment easily overlooks the matching between the component process details and repair requirements, resulting in process mismatch, production rework, and product delivery delay rate of more than 15%. (4) Weak adaptability: When faced with order additions, material supply delays or equipment failures, the entire process plan needs to be manually adjusted, which is time-consuming and can easily trigger a chain reaction, making it unable to adapt to the dynamic needs of multi-variety production; (5) Insufficient visualization and traceability: The scheduling results are mostly presented in the form of tables, lacking an intuitive process visualization interface, making it difficult to monitor the production progress in real time, and the reasons for scheduling deviations cannot be quickly traced. Summary of the Invention

[0004] This invention aims to address the technical problems of low production scheduling efficiency, unreasonable resource allocation, poor accuracy, weak responsiveness, and insufficient visualization and traceability of multiple types of rail vehicles in the prior art, and provides a smart optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A method for intelligent optimization and automatic scheduling of multiple types of rail vehicle products based on the SMART system includes the following steps: Step S1: Modeling scheduling basic data based on the SMART system; Build a standardized data model covering site, production line, personnel, components, and orders to provide accurate data support for scheduling: Step S2: Construction of a multi-constraint intelligent scheduling model; Based on the core constraints of rail vehicle production, a multi-constraint intelligent scheduling model adapted to the production of multiple varieties is constructed to ensure the feasibility and optimization of the scheduling scheme; Step S3: Improve the scheduling optimization solution of the genetic algorithm; An improved genetic algorithm is used to solve the multi-constraint intelligent scheduling model, and the optimal scheduling scheme is generated through iterative optimization. Step S4: Dynamic adjustment and real-time monitoring mechanism; Based on the real-time data interaction capabilities of the SMART system, a dynamic adjustment mechanism for scheduling schemes is established to respond to various changes in the production process; Step S5: Visual human-computer interaction interface design; Design a human-computer interaction interface to support visual control and manual intervention throughout the entire scheduling process; Step S6: Work order issuance and execution closed loop; The optimized scheduling plan is transformed into executable production work orders, realizing closed-loop management of scheduling and execution.

[0006] In the above technical solution, the scheduling basic data modeling in step S1 specifically includes: (1) Site resource modeling; Clearly define the capacity limits and time constraints for each site; (2) Production line configuration modeling; Record the number, priority, and active status of production lines corresponding to each site, and establish a mapping between sites and production lines; (3) Component process modeling; Clearly define the component codes, names, models, repair requirements, and corresponding project codes; and clarify the standard operating procedures and process dependencies for each component. (4) Order demand modeling; Enter order information, break it down into executable process tasks, and set time thresholds and quality requirements for each process.

[0007] In the above technical solution, the construction of the multi-constraint intelligent scheduling model in step S2 specifically includes: (1) Core constraints: Site constraints: Strictly match site type with product model, and repair level with site repair level; Production line constraints: Based on production line priority and capacity limit, avoid overloading of a single production line; Time constraints: Follow the work schedule and rest schedule to ensure that the cycle time of the process matches the actual production situation; Process constraints: Ensure that the sequence of processes is consistent with the repair requirements of the components, and avoid cross-process scheduling; (2) Model objective function - i.e., multi-constraint intelligent scheduling model: with the core objectives of maximizing equipment utilization, minimizing order delivery cycle, and minimizing process waiting time, a multi-objective optimization model objective function is constructed. The details are as follows: in, These are the weighting coefficients.

[0008] In the above technical solution, step S3 specifically includes: (1) Coding design: The four-dimensional coding method of order-part-production line-time is adopted. The chromosome segment includes the order number, part code, assigned production line number, and process start / end time to ensure that the coding corresponds one-to-one with the actual production. (2) Population initialization: Based on the site and production line configuration, 50-100 initial scheduling schemes are randomly generated to ensure that the initial population covers different resource allocation combinations; (3) Fitness calculation: The objective function of the model in step S2 is used as the fitness function to calculate the fitness value of each scheduling scheme and screen the effective schemes that meet the constraints. (4) Genetic manipulation: Selection: Use roulette wheel selection to select the top 30% of individuals with the best fitness values ​​to enter the next generation, and retain the optimal solution; Crossover: A two-point crossover method is used to exchange production line allocation-time node segments of different chromosomes to ensure that process and time constraints are still met after crossover; Mutation: Randomly adjust the production line allocation or process time of some individuals with a probability of 0.05-0.1 to avoid the algorithm getting trapped in local optima; (5) Iteration termination: When the optimal fitness value fluctuates by less than 1% in 10 consecutive iterations, the iteration stops and the optimal scheduling scheme is output.

[0009] In the above technical solution, step S4 specifically includes: (1) Real-time data acquisition: Data from the production site is acquired through IoT sensors and SMART systems, including: Resource status: Equipment operating status (normal / fault / maintenance), personnel on-duty status; Production schedule: actual completion time of each process and component assembly progress; External changes: order changes, material supply delays, quality inspection results; (2) Deviation warning: Set a deviation threshold. When the collected data exceeds the threshold, the system will automatically trigger a warning and analyze the cause of the deviation. (3) Automatic adjustment: For different types of deviation, the improved genetic algorithm is invoked for rapid re-optimization: Equipment failure: Assign unfinished tasks from the faulty production line to an available production line in the same area; Material delays: Adjust the start time of relevant component processes and prioritize tasks with sufficient materials; Order changes: Update the order data model and regenerate the scheduling plan to ensure that the resource allocation of new orders does not conflict with that of existing orders; (4) Adjustment and synchronization: The adjusted scheduling plan is synchronized to the production site terminal and management personnel interface in real time, and the work order information is automatically updated.

[0010] In the above technical solution, the human-computer interaction interface in step S5 includes: (1) Parameter configuration module: Supports inputting / modifying site plan, production line configuration, rest time, component process flow basic data, and customizing scheduling weight coefficients and deviation thresholds; (2) Scheduling visualization module: Displays scheduling results, showing the process arrangement, start / end time, and cycle time of each component on different production lines. It supports filtering and viewing by production line, order, and component type. (3) Adjustment Operation Module: Provides two adjustment methods: Manual fine-tuning: Supports dragging and dropping process time nodes and modifying production line allocation, with subsequent related processes automatically adjusting accordingly; Batch adjustment: Supports modification of daily get off work hours and rest periods; the system automatically adapts to the time schedule of all related processes. (4) Data traceability module: records the generation log, adjustment record, and production execution data of the scheduling plan. It supports querying the entire process information by order number and component code, and generates a scheduling effect analysis report.

[0011] In the above technical solution, step S6 specifically includes: (1) Work order generation: Automatically generate production work orders based on the scheduling plan; (2) Work order issuance: The work order is pushed to the corresponding production line operation terminal through the SMART system, and the operator carries out the work according to the work order requirements; (3) Execution feedback: After the worker completes the process, he enters the actual completion time and quality status on the terminal. The system updates the production progress in real time, forming an execution closed loop.

[0012] The intelligent optimization and automatic scheduling method for multiple types of rail vehicle products based on the SMART system of the present invention has the following beneficial effects: (1) Improve scheduling efficiency: shorten the scheduling cycle of single batch multi-variety orders from 16 hours to less than 4 hours, and realize the rapid and accurate matching of processes and production lines; (2) Optimize resource allocation: Combine constraints such as site, production line, and personnel to increase the utilization rate of special equipment to over 85% and reduce production waiting time and resource idleness; (3) Ensure the accuracy of scheduling: Strictly follow the component process flow and repair requirements to reduce the process mismatch rate to below 3% and reduce rework costs; (4) Enhance dynamic response capability: Respond to sudden production situations within 10 minutes, automatically generate adjustment plans, and ensure that the order delivery delay rate is reduced to below 5%; (5) Achieve visualized management and control: Provide an intuitive scheduling visualization interface and work order issuance function, support real-time monitoring of production progress and data traceability, and assist in scientific decision-making. Attached Figure Description

[0013] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0014] Figure 1 This is a flowchart illustrating the steps of the intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system of the present invention. Detailed Implementation

[0015] The inventive concept of this invention is as follows: The present invention provides a method for intelligent optimization and automatic scheduling of multiple types of rail vehicle products based on the SMART system. It constructs a full-process scheduling system of "data modeling - intelligent optimization - dynamic adjustment - visual control". Based on the core idea of ​​the SMART system and combined with the production characteristics of core components of rail vehicles, it realizes intelligent optimization and automatic scheduling of multi-type production.

[0016] The present invention will now be described in detail with reference to the accompanying drawings.

[0017] like Figure 1 As shown, the intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system of the present invention includes the following steps: Step S1: Modeling scheduling basic data based on the SMART system; Build a standardized data model covering site, production line, personnel, components, and orders to provide accurate data support for scheduling: (1) Site resource modeling: Enter the information in Table 1 "Site Plan Configuration", including production line group (such as bogie business unit maintenance production line group, assembly production line group), site type (five-type vehicle installation, three-type vehicle installation), repair level matching relationship, working hours (08:10-17:00) and rest configuration (Table 2: 12:00-13:00 lunch break), and clarify the capacity limit and time constraints of each site; Table 1: Site Plan Configuration

[0018] Table 2: Venue Rest Area Configuration

[0019] (2) Production line configuration modeling: Associate Table 3 "Site production line configuration" to record the number of production lines corresponding to each site (e.g., the bogie business unit assembly production line group includes production line 1 and production line 2), priority and effective status, and establish the association mapping between sites and production lines; Table 3: Site Production Line Configuration

[0020] (3) Component process modeling: Import Table 4 “Component process flow details”, clarify the code, name, model, repair level requirements and corresponding project code of each component (such as component M000003724761 corresponding to repair level “five-level repair one round”, project code A-3506), sort out the standard operating procedures and process dependencies of each component; Table 4: Detailed Process Flow of Components

[0021] (4) Order demand modeling: Enter order information, including product model, repair level requirements, delivery cycle, and parts list. Break it down into executable process tasks according to the "clarification" principle of the SMART system, and set the time threshold and quality requirements of each process according to the "quantification" principle.

[0022] Step S2: Construction of a multi-constraint intelligent scheduling model; Based on the core constraints of rail vehicle production, a multi-constraint intelligent scheduling model is constructed to adapt to the production of multiple product types, ensuring the feasibility and optimization of the scheduling scheme: (1) Core constraints: Site constraints: Strict matching of site type and product model (e.g., a site for installing Type 5 vehicles can only undertake the production of Type 5 vehicle parts), and compatibility of repair level and site repair level (e.g., parts for Level 4 repair can only be assigned to sites with Level 4 repair qualifications). Production line constraints: Based on production line priority and capacity limit, avoid overloading of a single production line (such as the balanced allocation of tasks between production lines 1 and 2 with priority 1 in Table 3). Time constraints: Follow the work hours and rest schedule (no work is scheduled from 12:00 to 13:00) to ensure that the work process rhythm matches the actual production (as shown in Table 5, each process avoids the lunch break period); Table 5: System Automatic Scheduling and Adjustment

[0023] Process constraints: In accordance with the requirements of Table 6 "Component Process Flow Details", ensure that the sequence of processes is consistent with the component repair level requirements and that there are no cross-process scheduling issues.

[0024] Table 6: Detailed Process Flow of Components

[0025] (2) Model objective function - i.e., multi-constraint intelligent scheduling model: with the core objectives of "maximizing equipment utilization, minimizing order delivery cycle, and minimizing process waiting time", a multi-objective optimization model objective function is constructed. The formula is as follows: in, These are weighting coefficients (default values ​​are 0.4, 0.3, and 0.3 respectively), which can be dynamically adjusted based on order priority.

[0026] Step S3: Improve the scheduling optimization solution of the genetic algorithm; An improved genetic algorithm is used to solve the multi-constraint intelligent scheduling model, and the optimal scheduling scheme is generated through iterative optimization. (1) Coding design: The four-dimensional coding method of "order-part-production line-time" is adopted. The chromosome segment includes the order number, part code, assigned production line number, and process start / end time to ensure that the coding corresponds one-to-one with the actual production. (2) Population initialization: Based on the site and production line configuration, 50-100 initial scheduling schemes are randomly generated to ensure that the initial population covers different resource allocation combinations; (3) Fitness calculation: The objective function of the model in step S2 is used as the fitness function to calculate the fitness value of each scheduling scheme and screen the effective schemes that meet the constraints. (4) Genetic manipulation: Selection: Use roulette wheel selection to select the top 30% of individuals with the best fitness values ​​to enter the next generation, and retain the optimal solution; Crossover: A two-point crossover method is used to exchange the "production line allocation-time node" segments of different chromosomes to ensure that the process and time constraints are still met after the crossover. Mutation: Randomly adjust the production line allocation or process time of some individuals with a probability of 0.05-0.1 to avoid the algorithm getting trapped in local optima; (5) Iteration termination: When the optimal fitness value fluctuates by less than 1% in 10 consecutive iterations, the iteration stops and the optimal scheduling scheme is output (as shown in Table 5 for process allocation and time plan).

[0027] Step S4: Dynamic adjustment and real-time monitoring mechanism; Based on the real-time data interaction capabilities of the SMART system, a dynamic adjustment mechanism for scheduling schemes is established to respond to various changes in the production process: (1) Real-time data acquisition: Data from the production site is acquired through IoT sensors and SMART systems, including: Resource status: Equipment operating status (normal / fault / maintenance), personnel on-duty status; Production schedule: actual completion time of each process and component assembly progress; External changes: order modifications (additions / cancellations), material supply delays, quality inspection results; (2) Deviation warning: Set deviation thresholds (such as process completion time deviation ≥ 30 minutes, material delay ≥ 2 hours). When the collected data exceeds the threshold, the system will automatically trigger a warning and analyze the cause of the deviation. (3) Automatic adjustment: For different types of deviation, the improved genetic algorithm is invoked for rapid re-optimization: Equipment failure: Assign unfinished tasks from the faulty production line to an available production line in the same area; Material delays: Adjust the start time of relevant component processes and prioritize tasks with sufficient materials; Order changes: Update the order data model and regenerate the scheduling plan to ensure that the resource allocation of new orders does not conflict with that of existing orders; (4) Adjustment and synchronization: The adjusted scheduling plan is synchronized to the production site terminal and management personnel interface in real time, and the work order information is automatically updated.

[0028] Step S5: Visual human-computer interaction interface design; Design a professional human-computer interaction interface to support visualized management and manual intervention throughout the scheduling process: (1) Parameter configuration module: Supports inputting / modifying basic data such as site plan, production line configuration, rest time, and component process flow, and allows customization of scheduling weight coefficients and deviation thresholds; (2) Scheduling visualization module: The scheduling results are displayed in the form of Gantt chart, clearly showing the process arrangement, start / end time and cycle time of each component on different production lines. It supports filtering and viewing by production line, order and component type. (3) Adjustment Operation Module: Provides two adjustment methods: Manual fine-tuning: Supports dragging and dropping process time nodes and modifying production line allocation, with subsequent related processes automatically adjusting accordingly; Batch adjustment: Supports modification of daily get off work hours and rest periods; the system automatically adapts to the time schedule of all related processes. (4) Data traceability module: records the generation log, adjustment record, and production execution data of the scheduling plan. It supports querying the entire process information by order number and component code, and generates a scheduling effect analysis report.

[0029] Step S6: Work order issuance and execution closed loop; The optimized scheduling plan is transformed into executable production work orders, achieving closed-loop management of scheduling and execution: (1) Work order generation: The system automatically generates production work orders based on the scheduling plan, including work order number, component information, allocated production line, process requirements, time nodes, quality standards, etc. (2) Work order issuance: The work order is pushed to the corresponding production line operation terminal through the SMART system, and the operator carries out the work according to the work order requirements; (3) Execution feedback: After the worker completes the process, he enters the actual completion time and quality status on the terminal. The system updates the production progress in real time, forming an execution closed loop.

[0030] The specific implementation examples of the intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system of the present invention are as follows: 1. Implementation scenarios; Taking the bogie division as an example, it undertakes the maintenance and assembly of bogies for Type III and Type V trains, involving various repair levels such as Level IV and Level V. The production resource allocation is as follows: Site and production lines: 2 production line groups (bogie business unit maintenance production line group and assembly production line group), with a total of 3 operational production lines (1 maintenance production line group and 2 assembly production lines group). Working hours: 08:10—17:00, lunch break 12:00—13:00; Order requirements: Bulk orders (50 bogies for Type 5 trains undergoing Level 5 overhaul, with a delivery period of 30 days), Custom orders (10 bogies for Type 3 trains undergoing Level 4 overhaul, with a delivery period of 15 days); Component types: including 8 core components such as EMU bogies and trailer bogies (as shown in Table 6), and the repair level of each component corresponds one-to-one with the project code.

[0031] 2. Implementation steps; Step 1: Basic data entry and modeling; Enter the following data through the SMART system: (1) Site planning configuration: Import site configuration information and clarify the vehicle type compatibility (Type 5 vehicles / Type 3 vehicles) and repair range of each site; (2) Production line configuration: Import production line information and set the priority to 1 for all production lines; (3) Rest arrangement: Import lunch break rules (12:00~13:00); (4) Component process flow: Import the component details and clarify the repair requirements of each component and the relationship between the project code; (5) Order data: Enter the product quantity, delivery cycle, and parts list of the order, which are broken down into 120 specific processes (each bogie contains 12 core processes).

[0032] Step 2: Construction and optimization of multi-constraint intelligent scheduling model; (1) Set constraints: Clearly define that bogies for Type 5 cars are only assigned to the “Type 5 car installation” site (e.g., project codes A-2501, A-2502), and bogies for Type 3 cars are only assigned to the “Type 3 car installation” site (e.g., A-3505, A-3506). (2) Configure objective function weights: Prioritize delivery time for customized orders (Type III vehicle, Level IV repair), and set... Bulk orders (five types of vehicles, five levels of repair) will be given priority in ensuring production capacity. ; (3) Algorithm operation: Start the improved genetic algorithm, set the population size to 80, the number of iterations to 120, and finally output the optimal scheduling scheme (as shown in Table 5). (4) Maintenance production line group: Responsible for the fifth-level maintenance of 20 bogies of Type 5 cars, with the process time being 8:00~17:00 (avoiding lunch break); (5) Assembly production line group production line 1: undertakes the task of 20 bogies for type 5 cars + 5 bogies for type 3 cars, and is scheduled in the order of "type 5 cars → type 3 cars"; (6) Assembly production line group production line 2: undertakes the task of 10 bogies for type 5 vehicles + 5 bogies for type 3 vehicles, and ensures that customized orders are delivered within 15 days.

[0033] Step 3: Production Execution and Dynamic Adjustment; (1) Work order issuance: The system generates 120 work orders according to the scheduling plan and pushes them to the operation terminals of each production line; (2) Handling of emergencies: On the 10th day of production, the assembly line production line 1 stopped due to equipment failure. The system collected fault information in real time and triggered an early warning. The adjustment was completed within 10 minutes: the unfinished task of 8 bogies for Type 5 vehicles on production line 1 was assigned to production line 2; the process time plan of production line 2 was adjusted, and some processes were extended to the next day to ensure that the delivery time of customized orders was not affected. (3) Progress monitoring: Managers can view the progress of each production line in real time through the visual interface. If they find that the progress of a certain process in production line 2 is 15 minutes behind, they can manually adjust the start time of the subsequent process and the system will automatically update.

[0034] 3. Verification of implementation results; After applying the method of this invention, production indicators were significantly optimized, as shown in Table 7: Table 7 Comparison of Production Indicators

[0035] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for intelligent optimization and automatic scheduling of multiple types of rail vehicle products based on a SMART system, characterized in that, Includes the following steps: Step S1: Modeling scheduling basic data based on the SMART system; Build a standardized data model covering site, production line, personnel, components, and orders to provide accurate data support for scheduling: Step S2: Construction of a multi-constraint intelligent scheduling model; Based on the core constraints of rail vehicle production, a multi-constraint intelligent scheduling model adapted to the production of multiple product types is constructed to ensure the feasibility and optimization of the scheduling scheme; Step S3: Improve the scheduling optimization solution of the genetic algorithm; An improved genetic algorithm is used to solve the multi-constraint intelligent scheduling model, and the optimal scheduling scheme is generated through iterative optimization. Step S4: Dynamic adjustment and real-time monitoring mechanism; Based on the real-time data interaction capabilities of the SMART system, a dynamic adjustment mechanism for scheduling schemes is established to respond to various changes in the production process; Step S5: Visual human-computer interaction interface design; Design a human-computer interaction interface to support visualized management and manual intervention throughout the entire scheduling process; Step S6: Work order issuance and execution closed loop; The optimized scheduling plan is transformed into executable production work orders, realizing closed-loop management of scheduling and execution.

2. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 1, characterized in that, The scheduling basic data modeling in step S1 specifically includes: (1) Site resource modeling; Clearly define the capacity limits and time constraints for each site; (2) Production line configuration modeling; Record the number, priority, and active status of production lines corresponding to each site, and establish a mapping between sites and production lines; (3) Component process modeling; Clearly define the component codes, names, models, repair requirements, and corresponding project codes; and clarify the standard operating procedures and process dependencies for each component. (4) Order demand modeling; Enter order information, break it down into executable process tasks, and set time thresholds and quality requirements for each process.

3. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 1, characterized in that, The construction of the multi-constraint intelligent scheduling model in step S2 specifically includes: (1) Core constraints: Site constraints: Strictly match site type with product model, and repair level with site repair level; Production line constraints: Based on production line priority and capacity limit, avoid overloading of a single production line; Time constraints: Follow the work schedule and rest schedule to ensure that the cycle time of the process matches the actual production situation; Process constraints: Ensure that the sequence of processes is consistent with the repair requirements of the parts, and avoid cross-process scheduling; (2) Model objective function - i.e., multi-constraint intelligent scheduling model: with the core objectives of maximizing equipment utilization, minimizing order delivery cycle, and minimizing process waiting time, a multi-objective optimization model objective function is constructed. The details are as follows: in, These are the weighting coefficients.

4. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 3, characterized in that, Step S3 specifically includes: (1) Coding design: The four-dimensional coding method of order-part-production line-time is adopted. The chromosome segment includes the order number, part code, assigned production line number, and process start / end time to ensure that the coding corresponds one-to-one with the actual production. (2) Population initialization: Based on the site and production line configuration, 50-100 initial scheduling schemes are randomly generated to ensure that the initial population covers different resource allocation combinations; (3) Fitness calculation: The objective function of the model in step S2 is used as the fitness function to calculate the fitness value of each scheduling scheme and screen the effective schemes that meet the constraints. (4) Genetic manipulation: Selection: Use roulette wheel selection to select the top 30% of individuals with the best fitness values ​​to enter the next generation, and retain the optimal solution; Crossover: A two-point crossover method is used to exchange production line allocation-time node segments of different chromosomes to ensure that process and time constraints are still met after crossover; Mutation: Randomly adjust the production line allocation or process time of some individuals with a probability of 0.05-0.1 to avoid the algorithm getting trapped in local optima; (5) Iteration termination: When the optimal fitness value fluctuates by less than 1% in 10 consecutive iterations, the iteration stops and the optimal scheduling scheme is output.

5. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 1, characterized in that, Step S4 specifically includes: (1) Real-time data acquisition: Data from the production site is acquired through IoT sensors and SMART systems, including: Resource status: Equipment operating status: normal / fault / maintenance; personnel on-site status; Production schedule: actual completion time of each process and component assembly progress; External changes: order changes, material supply delays, quality inspection results; (2) Deviation warning: Set a deviation threshold. When the collected data exceeds the threshold, the system will automatically trigger a warning and analyze the cause of the deviation. (3) Automatic adjustment: For different types of deviation, the improved genetic algorithm is invoked for rapid re-optimization: Equipment failure: Assign unfinished tasks from the faulty production line to an available production line in the same area; Material delays: Adjust the start time of relevant component processes and prioritize tasks with sufficient materials; Order changes: Update the order data model and regenerate the scheduling plan to ensure that the resource allocation of new orders does not conflict with that of existing orders; (4) Adjustment and synchronization: The adjusted scheduling plan is synchronized to the production site terminal and management personnel interface in real time, and the work order information is automatically updated.

6. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 1, characterized in that, The human-computer interaction interface in step S5 includes: (1) Parameter configuration module: Supports inputting / modifying basic data such as site plan, production line configuration, rest time, and component process flow, and allows for custom scheduling weight coefficients and deviation thresholds; (2) Scheduling visualization module: Displays scheduling results, showing the process arrangement, start / end time, and cycle time of each component on different production lines. It supports filtering and viewing by production line, order, and component type. (3) Adjustment Operation Module: Provides two adjustment methods: Manual fine-tuning: Supports dragging and dropping process time nodes and modifying production line allocation, with subsequent related processes automatically adjusting accordingly; Batch adjustment: Supports modification of daily get off work hours and rest periods; the system automatically adapts to the time schedule of all related processes. (4) Data traceability module: records the generation log, adjustment record, and production execution data of the scheduling plan. It supports querying the entire process information by order number and component code, and generates a scheduling effect analysis report.

7. The intelligent optimization automatic scheduling method for multiple types of rail vehicle products based on the SMART system according to claim 1, characterized in that, Step S6 specifically includes: (1) Work order generation: Automatically generate production work orders based on the scheduling plan; (2) Work order issuance: The work order is pushed to the corresponding production line operation terminal through the SMART system, and the operator carries out the work according to the work order requirements; (3) Execution feedback: After the worker completes the process, he enters the actual completion time and quality status on the terminal. The system updates the production progress in real time, forming an execution closed loop.

Images