Intelligent production scheduling method for automobile electronics manufacturing

Abstract

The present application belongs to the technical field of intelligent manufacturing, and specifically relates to an intelligent production scheduling method for automobile electronic manufacturing, comprising the following steps: step S100: constructing a production scheduling model based on automobile electronic multi-stage production; step S200: dividing multiple constraint conditions of automobile electronic production into different priorities, and constructing a target function based on the priorities; step S300: solving the model by adopting a step-by-step genetic algorithm; and step S400: outputting a final iteration-optimized production scheduling scheme; the present application, on one hand, establishes a unique mathematical model through deep understanding of an actual automobile production scene, and better reflects constraint requirements on order sequences in an actual automobile electronic manufacturing scene; on the other hand, through a step-by-step processing scheme, specific constraint priorities and actual production influences are satisfied one by one, and the solving speed and solving quality are greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology, specifically relating to an intelligent scheduling method for automotive electronics manufacturing. Background Technology

[0002] In recent years, the automotive electronics manufacturing industry has ushered in unprecedented development opportunities globally. As a product of the deep integration of electronic information technology and automotive manufacturing, automotive electronics has evolved from its initial auxiliary functions (such as electronic ignition and radio) to a core component determining the performance, safety, comfort, and intelligence levels of modern vehicles. The accelerated advancement of automotive intelligence and connectivity has led to a surge in demand for complex electronic systems such as Advanced Driver Assistance Systems (ADAS), in-vehicle infotainment systems, and autonomous driving technologies. Secondly, the explosive growth of new energy vehicles, with their high dependence on core electronic components such as battery management systems (BMS), motor controllers, and electronic control systems, has further increased the value of electronics per vehicle. Against this backdrop, the global industrial chain has continuously improved, forming a competitive landscape dominated by international giants such as Bosch, Continental, and Denso, while domestic companies such as Huayang Group and Desay SV are rapidly emerging. However, despite the continuous improvement in market size and technological level, the industry has also exposed many challenges, among which insufficient refinement and customization capabilities in production management have become a key bottleneck restricting the improvement of enterprise competitiveness.

[0003] For the scheduling problem of automotive electronics manufacturing, many companies have not developed customized scheduling algorithms and rely excessively on industrial solvers. However, there are many intelligent scheduling methods worth learning from in similar production companies. For example, Reference [1] discloses a lean intelligent scheduling algorithm. In the study of the scheduling problem of a mixed flow shop with ordered related tool change time and limited buffer between the order, the algorithm comprehensively considers the tool change constraint and the buffer between the order constraint, and uses the second-by-second simulation method to solve the problem, which verifies the practicality and effectiveness of the proposed method. Reference [2] discloses a reconfigurable production line integrated scheduling method based on genetic algorithm. In view of the dynamic production needs of intelligent manufacturing workshop, it comprehensively considers the workpiece processing needs, production line type, equipment availability and process constraints, and constructs an integrated production line construction and workpiece scheduling optimization model with the goal of minimizing the maximum completion time. The method innovatively combines heuristic greedy algorithm and genetic algorithm. The former dynamically constructs and reconstructs production line instances based on equipment availability and average theoretical processing time, while the latter realizes the global optimization of workpiece allocation and scheduling through two-stage encoding and customized cross mutation operation. Under the premise of satisfying multiple constraints, this method significantly improves the utilization rate of the production line through efficient equipment allocation, enhances the flexibility of the production system, and shortens the production cycle. It is particularly suitable for intelligent manufacturing scenarios with multiple varieties and small batches, and provides an efficient and robust scheduling solution for complex production environments. Reference [3] solves the target scheduling and warehouse logistics scheduling problems through a multi-vehicle scheduling method based on genetic algorithm. An improved repair method is used to generate the initial solution; with the total time as the target, the parent generation is selected by roulette wheel selection based on fitness for crossover and mutation operations to generate offspring; the improved repair in the initial solution is repeated for the offspring; the parent generation and offspring generation are merged, and individuals with high fitness are retained to enter the next generation; with the conflict probability as the target, the multi-vehicle trajectory planning is optimized by timestamp prediction and artificial potential field; the conflict cost and total distance are used as optimization targets, and the non-dominated solution is obtained by non-dominated sorting genetic algorithm; the time and conflict cost are normalized, the Euclidean distance between each solution and the ideal point is calculated, and the solution with the smallest distance is selected as the optimal solution. The advantages are: multi-objective optimization is achieved through non-dominated sorting and crowding distance calculation; while considering conflict risk in global planning, a reasonable route is planned. Reference [4] discloses a scheduling system based on MOSO encoding genetic algorithm, including: a computing power execution module: used to split the scheduling task into a set of sub-tasks based on machine / Lot dimension, and process them through a multi-threaded parallel computing architecture, and output a chromosome population containing local optimal solutions; a MOSO encoding module: used to map the chromosome population into a structured gene sequence containing machine code and opcode using a multi-structure optimization encoding mechanism, wherein the machine code represents Lot. The machine allocation relationship is defined, with the opcode representing the Lot processing order. By breaking down large-scale scheduling tasks into a set of sub-tasks at the machine / Lot dimension, a multi-threaded parallel computing architecture is used to achieve distributed task processing, shortening the computation time of complex scheduling problems. A structured mapping between machine code and opcode expresses the Lot... Machine allocation relationship and processing sequence. Reference [5] discloses a UUV cluster dynamic task planning method, program, equipment and storage medium based on multi-objective genetic algorithm, which belongs to the field of underwater multi-UUV cooperative detection of multiple targets. The method first uses sequential encoding to complete chromosome representation for the path planning of regional task points, and selects and randomly sorts the chromosomes, selecting the chromosome with the highest fitness value from each group; then, in each group of the current population, the optimal parent individual is selected for crossover and mutation operations to improve the quality of offspring individuals, and the next generation population is updated; finally, multiple optimal individuals are obtained, and the task planning path results are output. The search area is divided into multiple task sub-regions, and the dynamic task planning of the UUV cluster is completed according to the target distribution of the sub-regions and the situation of each UUV. The method of the present invention can adjust the search strategy in real time according to the real-time detection results and environmental changes, thereby effectively alleviating the non-uniformity of target distribution and the difference in cluster efficiency. Reference [6] discloses a multi-capacity AGV task scheduling method based on an improved genetic algorithm and a hierarchical clustering algorithm, including: first, using the initial solution generation method of the hierarchical clustering algorithm, clustering the task points, performing double-layer encoding between and within clusters, and randomly generating the initial delivery order; at the same time, improving the genetic algorithm by replacing the mutation operation in the genetic algorithm with a differential operation; then, iteratively updating the improved genetic algorithm and the double-layer encoding strategy; finally, changing the chromosome encoding method to a single-layer encoding method, continuing to improve the iteration of the genetic algorithm, and obtaining a multi-capacity AGV scheduling scheme. The multi-capacity unmanned tractor task scheduling method disclosed in this invention reduces the travel distance of the multi-capacity unmanned tractor and improves the timeliness of task completion. Reference [7] discloses an improved genetic algorithm for autonomous scheduling in cotton spinning workshops. The method includes: using four performance indicators—processing waiting time, processing load time, processing time, and repair time—as optimization objectives and establishing corresponding objective functions; using processing time and processing time as boundary conditions to form constraint conditions; and using encoding and decoding operations in the genetic algorithm to realize the conversion between individuals in the population and scheduling schemes. This invention uses optimization objectives to establish a multi-objective model and improves the traditional genetic algorithm by using crossover, mutation, and other operations. In the process of solving the algorithm, as the number of iterations increases, the change in the population mean tends to stabilize, and the change in the population mean also tends to stabilize. Through simulation test cases, the performance of the algorithm is compared and analyzed, and the superiority and effectiveness of the improved genetic algorithm in solving the production scheduling problem in cotton spinning workshops are verified. Reference [8] discloses a heterogeneous UAV collaborative task allocation method and system based on genetic algorithm, which belongs to the field of heterogeneous UAV collaborative task allocation technology, including: using k Means clustering divides the target region into sub-regions. The first execution strategy of the heterogeneous UAV in each sub-region is generated by the improved genetic algorithm. The second execution strategy of the heterogeneous UAV in each sub-region is generated by the deep reinforcement learning network. The third execution strategy of the heterogeneous UAV in each sub-region is determined based on the high-frequency task existence information of each sub-region. The ranking penalty factor of each task is analyzed to determine the fourth execution strategy of the heterogeneous UAV in each sub-region and the fifth execution strategy of the heterogeneous UAV in each sub-region. The present invention effectively reduces the total flight distance of the system, improves the task scheduling efficiency, realizes the dynamic balance of UAV task resource load, and realizes the suppression of inter-task interference and intelligent coordination of resource conflicts, thereby improving the stability of task completion and the intelligence of scheduling. Reference [9] proposes a task scheduling method, medium and equipment for integrated avionics system based on the improved genetic algorithm, which belongs to the field of integrated avionics system task scheduling. The present invention addresses the characteristics of complex task scheduling and limited resources in integrated avionics system by constructing a task scheduling model from task to partition. The maximum partition usage index of task allocation and the average task completion time are taken as optimization objectives, and the constraints of task schedulability, memory, CPU utilization and security are comprehensively considered. The initial population is generated by chaotic mapping, and an improved multi-objective genetic algorithm is introduced. The population evolution is achieved by simulating binary crossover, integer mutation, constraint processing and other strategies, and the Pareto optimal solution set is output for the system to select the optimal scheduling scheme. This invention can effectively improve the feasibility of the scheduling scheme, the efficiency of resource utilization and the flexibility of scheduling, and is suitable for complex avionics system environments with multiple tasks and multiple constraints. Reference

[10] discloses a flexible flow shop scheduling method based on genetic algorithm that considers mold changing. The method constructs a multi-objective optimization model with the goal of minimizing the maximum completion time, reducing the delay cost, reducing the processing cost and optimizing the mold changing cost. On the basis of the traditional flexible flow shop scheduling problem with multiple products and multiple orders, it innovatively introduces the mold changing operation constraint and comprehensively considers the actual production constraints such as order release time and delivery period. The above related results are all production scheduling methods designed by intelligent optimization algorithms such as genetic algorithms in similar scheduling scenarios. These methods are all methods dedicated to specific scheduling scenarios. If they are directly applied to the automotive electronics manufacturing scenario, there will be problems such as model inapplicability and poor algorithm solution effect. Currently, most automotive electronics manufacturers still use manual scheduling or solver scheduling. The former is prone to errors, takes a long time, and can only ensure that hard constraints are met, but cannot correctly evaluate the satisfaction of soft constraints. The latter, although it can guarantee the constraint satisfaction rate, has the disadvantages of high deployment cost, poor performance due to method specialization, and generally long scheduling time.

[0004] To efficiently solve complex scheduling problems, this application first establishes a mixed-integer programming model. Considering the multi-objective nature of the problem, it also designs an efficient genetic algorithm framework, providing a practical solution for actual production scheduling, improving the efficiency of the solution, and enabling the solution to gradually approach the optimal solution with a relatively small number of iterations. The dedicated method for production scheduling in automotive electronics manufacturing proposed in this invention can obtain a production scheduling scheme with a high constraint satisfaction rate in a short time, and has low deployment costs.

[0005] The following is a detailed explanation of the references: [1] [Wang Jiangtao, 2020] A lean intelligent scheduling algorithm; [2] [Liu Zhaoyu, 2025] A reconfigurable production line integrated scheduling method based on genetic algorithm; [3] [Liu Weiping, 2025] A multi-vehicle scheduling method based on genetic algorithm; [4] [Xu Zhong, 2025] A scheduling system based on a gene genetic algorithm using MOSO encoding; [5] [Zhao Qiwei, 2025] A method, program, device and storage medium for dynamic task planning of UUV clusters based on multi-objective genetic algorithm; [6] [Zhan Yan, 2025] A multi-load AGV task scheduling method based on improved genetic algorithm and hierarchical clustering algorithm; [7] [Shao Jingfeng, 2025] An improved genetic algorithm for autonomous scheduling in cotton spinning workshops; [8] [Li Xiao, 2025] A method and system for collaborative task allocation of heterogeneous UAVs based on genetic algorithm; [9] [Feng Shuying, 2025] A method, medium and equipment for scheduling integrated avionics system based on improved genetic algorithm;

[10] [Pei Zhi, 2025] A flexible flow shop scheduling method based on genetic algorithm considering mold changing. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent scheduling method for automotive electronics manufacturing, in order to solve the problems existing in the background art.

[0007] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows: A smart scheduling method for automotive electronics manufacturing includes the following steps: Step S100: Construct a production scheduling model based on multi-stage production of automotive electronics. The production scheduling model defines the set, parameters and decision variables related to products, processes, production lines, resources and time. Step S200: Divide the multiple constraints of automotive electronics production into different priorities according to the importance of production scheduling, and construct an objective function based on the priorities of the constraints; Step S300: The model is solved using a stepwise genetic algorithm. First, an initial population that prioritizes satisfying high-priority constraints is generated. Then, iterative optimization is performed through genetic operations and directional search combined with constraint information. Step S400: Output the final iteratively optimized production scheduling plan, which includes the process scheduling sequence of each order on each level of production line.

[0008] As a further limitation of the technical solution of the present invention, the set defined in step S100 includes: Product set, process set, production line set, time unit set, date set, all workpiece set, workpiece set of product p, resource type set, safety stock set, and batch set; The production line set includes a chip programming line set, a PCB board production line set, and an assembly and packaging line set.

[0009] As a further limitation of the technical solution of the present invention, the parameters defined in step S100 include: Basic parameters: product Total order quantity, workpieces Quantity, production lines Previous process Production rate, transmission time, preset positive number, process Is it possible to do it on the production line? Production and products Is a process required? Production line flow relationship, workpiece Delivery time, products The final process, specified start time, specified finish time, production line availability, available resources, and process resource requirements; Inventory and batch parameters: Safety stock target value, beginning inventory, planned shipment quantity, economic batch size, small batch threshold, whether it is a small batch, whether it is a work-in-process order, and sub-order relationships.

[0010] As a further limitation of the technical solution of the present invention, the decision variables defined in step S100 include: Assignment and time series variables: Process start marker, process start time, process completion time, production line allocation marker, and process sequence marker; Resource and system variables: Production line usage markings, total time span, and capacity utilization rate; Inventory and batch variables: Inventory levels, production quantities, batch presence indicators, batch size, and changeover costs.

[0011] As a further limitation of the technical solution of the present invention, in step S200, the constraint condition of priority 0 (highest level) is as follows: Constraints include primary resource capacity, secondary resource capacity, process allocation, production line monopoly, process sequence, production line flow, processing time, and sub-order sequence.

[0012] As a further limitation of the technical solution of the present invention, the priority of other constraints in step S200 is as follows: Priority 1: Production scheduling constraints; Priority 2: Specify date constraints; Priority 3: Inventory level constraints, initial inventory constraints, minimum safety stock constraints, economic batch size constraints, and small batch concentration constraints; Priority 4: Material completeness constraint; Priority 5 and 6: Capacity utilization constraints; [1] Priority 7: Switch to minimum constraint; Priority 8: SMT batch constraints and adjacent scheduling constraints; Priority 9: Parallel process interval constraint; Priority 10: Order delivery date constraint.

[0013] As a further limitation of the technical solution of the present invention, step S300 specifically includes: S310: Initial population generation sub-step: An initial population is generated using a coarse-sorting strategy based on constraint priority. This strategy first satisfies the constraint with the highest priority. Orders are sorted and mapped to chromosome sequences according to the rules of prioritizing mass-produced products, prioritizing those with larger safety stock deviations, and prioritizing those with smaller absolute inventory quantities. The chromosomes are encoded in the order of order or product number. S320: Perform genetic iterative optimization on the population, in each generation: ① Use the roulette wheel selection operator to select parents based on individual fitness; ② Gene exchange is performed on the selected parent generation using a two-point crossover operator; ③ Apply a mutation operator to the crossover individuals, wherein the mutation is to randomly select a continuous gene segment in the chromosome and move it to the end of the sequence; ④ During the iteration process, for chromosome regions that do not meet specific constraints, targeted domain search based on parental gene information is used for adjustment; S330: Decoding and Evaluation Sub-steps: The optimized chromosome sequence is decoded into a specific production scheduling plan. The start time and completion time of each process are calculated based on the product process path and production line production rate. The fitness of the plan is evaluated using the objective function.

[0014] As a further limitation of the technical solution of the present invention, the coarse sorting strategy in step S310 is specifically as follows: ① Calculate the deviation between the current inventory of each mass-produced product and the minimum, standard, and maximum safety stock targets; ② Select products in ascending order of "first-level inventory deviation (difference from minimum safety stock)". If all deviations are positive, check the second-level and third-level inventory deviations in turn. ③ During the selection process, the availability of the target production line is monitored in real time; if it is unavailable, the current product is skipped. ④ If the inventory of all mass-produced products is sufficient, then new product or spare parts orders with a specified production date will be scheduled.

[0015] As a further limitation of the technical solution of the present invention, the targeted domain search in step S320 specifically includes: Based on the location information of genes in the parent chromosome that do not meet the constraints, the order scheduling sequence at the corresponding locations is specifically adjusted to directly repair the constraint conflict.

[0016] The present invention has the following beneficial effects: ① By establishing a mathematical model to express the three-stage process flow of automotive electronics manufacturing, this method more comprehensively and realistically reflects the actual production scheduling requirements. In actual production, enterprises need to strictly follow the production process flow. The solution sequences constructed by this method all strictly follow the production process. The resulting production scheduling scheme, while fully satisfying the hard constraints, mathematically evaluates the satisfaction of the soft constraints. The obtained solution sequences have a higher constraint satisfaction rate than existing methods, a faster solution speed, and lower economic costs. ② It is more conducive to balancing operational efficiency and production costs. In actual production activities, the two goals of production efficiency and production cost control often cannot be achieved at the same time. Enterprises need to find a balance between production speed and production cost, that is, they need to find a reasonable production method that takes into account both high production efficiency and low production cost. In this method model, the hard constraints in the automotive electronics production process are fully satisfied first. At the same time, the factors that have a greater impact on production speed and production cost are satisfied in order of priority. These factors are considered in a unified manner in the objective function of the mathematical model through linear weighting. The selected objective is more in line with the enterprise's needs for cost reduction and efficiency improvement. Attached Figure Description

[0017] The present invention can be further illustrated by the non-limiting embodiments given in the accompanying drawings.

[0018] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the automotive electronics production process of the present invention; Figure 3 This is a schematic diagram of the automotive electronics manufacturing process of the present invention; Figure 4 This is a schematic diagram of the algorithm flow of the present invention; Figure 5 This is a chart showing the results of the invention being applied to production scheduling at an automotive electronics company. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0020] Example 1: This embodiment demonstrates the algorithm design of the technical solution of the present invention: 1. Data Preprocessing Process logic processing: Read the process path information of each product from the production scheduling data, that is, the production line number required for each product in each workshop. For example, if the product number is A10001, the corresponding process is IC workshop lines 1, 2, 3, and 4, SMT workshop lines 1 and 2, and final assembly workshop line 1. After obtaining the corresponding process information of the product, its production speed can be predicted according to its process and factory production history information.

[0021] Order data processing: Read order information from production scheduling data, and categorize orders into three types: mass production, new products, and spare parts. For mass production products, calculate their safety stock days. Calculation logic: The safety stock can meet the production plan's required number of days. For example, if there are 500 units of product A remaining on a given day, and the daily planned shipment is 200 units, then the safety stock days are 2 days. Mass-produced products have three levels of safety stock days: minimum, standard, and maximum. New products and spare parts should consider production start and end dates. Constraint handling: Read the production scheduling constraint information from the production scheduling data and establish a constraint information matrix. Constraints in the automotive electronics production scenario can be divided into process constraints and production scheduling efficiency constraints. Process constraints are mostly hard constraints, such as process logic constraints, while production scheduling efficiency constraints are mostly soft constraints, such as the number of material switching times.

[0022] 2. Sorting Constraint Priority Ranking: Prioritize the constraints considered in production scheduling; First Sorting (Coarse Sorting): In the first round of sorting, the algorithm fills the orders of the initial matrix into the grouping matrix according to the constraint information matrix, and the filling order follows four priority principles: Constraints with higher priority are preferred; mass-produced products are preferred; the difference between safety stock and the same-level safety stock is greater is preferred, and they are satisfied in the order of minimum, standard, and maximum; for multiple attributes of the same constraint, those with fewer quantities are preferred. Through the first-round sorting, an initial population that meets most of the higher-priority constraints can be obtained.

[0023] For example: Under the condition of limited production capacity, there are mass-produced product A, new product B, and spare part product C. If they cannot be scheduled simultaneously, the mass-produced product is scheduled first; for mass-produced products A and B, if the difference between the inventory days of A and B and the minimum safety stock days is -1 day and -2 days respectively, then product B is scheduled first. If the difference from the minimum safety stock days is the same, then compare the standard and maximum safety stock days in sequence. If they are the same, the priorities of these two types of products are exactly the same, and the product with the smaller absolute inventory quantity is scheduled first.

[0024] Second Sorting (Fine Sorting): Use the genetic algorithm to adjust some constraints that are still not satisfied.

[0025] Population Initialization: Generate the initial population through the coarse sorting strategy by randomly considering several priority principles.

[0026] Encoding of Solutions: Sequence encoding is used, and a list of order numbers or product numbers is used as the solution. For example: [A10001, A10002,..., A10100]; when decoding, the specific scheduling situation is decoded through the processes of each product. For example, the process of A10001 is IC01 - 04, SMT01 - 02, and the general assembly line 01. The processing volume of each process on each line is determined by historical data and the given values of the factory. Subsequently, the start processing time and end processing time of each process in each workshop during the production process can be calculated according to the product quantity.

[0027] Selection Operator: This method uses roulette wheel selection, and better solution sequences have a higher selection probability.

[0028] Crossover Operator: This method uses two-point crossover to exchange the order of two positions in the sequence.

[0029] Mutation Operator: Randomly select a position in the current sequence, take it as the first position, and randomly select a sequence with a length of n (n < N / 50, N is the total length) and exchange it to the end of the sequence.

[0030] Example 2: This example is a practical case as follows: The data in the following examples were randomly generated using sample data provided by an automotive electronics company as a template. Test data were generated for 20 products (mass production: new products: spare parts = 6:2:2, production line numbers: 10, 6, 4, quantity of each product ranging from 500, 600, 700, to 800, and the number of products in the example was the same), 100 products (mass production: new products: spare parts = 7:2:1, production line numbers: 20, 10, 6), 300 products (mass production: new products: spare parts = 8:1:1, production line numbers: 40, 20, 10), and 1000 products (mass production: new products: spare parts = 8:1:1, production line numbers: 100, 60, 40). The production speed of all products in each process was 100-150 units / hour. Method testing was conducted in a PTYHON 3.8 environment on a workstation equipped with an Intel(R) Core(TM) i7-14700k processor, 32GB of memory, and an Intel(R) UHDGraphics 770.

[0031] Specifically, consider the following constraints: Process constraints: Main resource constraint: When the resources of a certain production line are not in operation, the production line will not be used, and the production line will not be scheduled to produce during the time when the resources are not in production; Material constraints: Production scheduling is based on the process route sequence of the material codes; for sub-orders under the same order, production scheduling needs to be based on the material sequence: the sub-order at the upper level (such as final assembly) supplies the sub-order at the lower level (such as SMT); Production constraints: Work-in-process production scheduling constraints: For work-in-process orders, production scheduling is fixed based on the actual start time; Specified date constraint: For new product or spare parts orders with a specified date, production will be scheduled according to the specified date. Inventory level constraints: From the perspective of production line / capacity, meet the minimum safety stock → standard safety stock → maximum safety stock of finished products; Final assembly economic batch constraint: Based on the economic batch size of finished products, production should be scheduled as close as possible to the inventory level when considering inventory levels; Small-batch centralized constraints: For new products and spare parts, production is scheduled in a concentrated manner on the first appearance date of the current month; Maximum capacity utilization: Considering the overall situation, capacity utilization = planned time ÷ available capacity time; Changeover constraint: Minimize the number of changes for material codes scheduled on the same production line (equipment); Minimize the off-line interval of SMT parallel processes: Identify whether a process has multiple preceding processes and constrain the off-line interval of all preceding processes to be as small as possible.

[0032] Production scheduling method: Step 1: Calculate the fixed production time for each product based on its process, i.e. the time spent processing each step, and calculate the number of sub-orders in each workshop (some products require the production of multiple sub-orders in certain steps); Step 2: Calculate the current safety stock days for each mass-produced product, and calculate the difference with the level 3 safety stock. Divide the difference into level 1, level 2, and level 3 inventory deviations and sort them in ascending order. Step 3: Select products for production scheduling one by one according to the first-level inventory deviation. When selecting, consider the product process and check if the production line is available. If it is not available, skip it and check it again the next day. If all products and inventory deviations are greater than 0, that is, they all meet the minimum safety stock, check the available production lines for the day. If there are available production lines, check if there are any orders with a specified production date (plus or minus one day). If there are, schedule the new products or spare parts orders with the specified production date, and schedule the remaining new products and spare parts orders for the near future (except for those with specified dates). If not, schedule the orders according to the second-level and third-level inventory deviations in turn. Step 4: Check whether the current production schedule meets the final assembly economic batch requirement (i.e., the quantity of a certain type of material required for the current production product does not meet the quantity requirement). If it does not meet the requirement, exchange orders and select the order with the smallest inventory deviation that meets the final assembly economic batch requirement. Step 5: Calculate capacity utilization, number of switching times, and time interval between offline operations; select products for exchange based on inventory deviation using a genetic algorithm exchange strategy to satisfy these three constraints.

[0033] For detailed production schedule results, please refer to the following: Figure 5 As shown in the chart, it can be seen that the technical solution of the present invention has a high capacity utilization rate when outputting a production scheduling scheme that meets the constraints for a product quantity of 20-1000, and the production scheduling calculation time is less than 1 minute, which is timely in practical application.

[0034] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A smart scheduling method for automotive electronics manufacturing, characterized in that: Includes the following steps: Step S100: Construct a production scheduling model based on multi-stage production of automotive electronics. The production scheduling model defines the set, parameters and decision variables related to products, processes, production lines, resources and time. Step S200: Divide the multiple constraints of automotive electronics production into different priorities according to the importance of production scheduling, and construct an objective function based on the priorities of the constraints; Step S300: The model is solved using a stepwise genetic algorithm. First, an initial population that prioritizes satisfying high-priority constraints is generated. Then, iterative optimization is performed through genetic operations and directional search combined with constraint information. Step S400: Output the final iteratively optimized production scheduling plan, which includes the process scheduling sequence of each order on each level of production line.

2. The intelligent scheduling method for automotive electronics manufacturing according to claim 1, characterized in that: The set defined in step S100 includes: Product Collection ( ), process set ( Production line assembly ( Time unit set ( Date collection ( ), collection of all workpieces ( The set of workpieces for product p ( ), resource type set ( Safety Stock Set ( and batch sets ; The production line set Includes chip programming cable set PCB board production line assembly and assembly and packaging line collection .

3. The intelligent scheduling method for automotive electronics manufacturing according to claim 2, characterized in that: The parameters defined in step S100 include: Basic parameters: Representative products Total number of orders; Representative workpiece Quantity; Representative production line Previous process ; production rate; Represents transmission time; Represents a pre-defined positive number; Representative process Is it possible to do it on the production line? Up to production; Representative products Is a process required? ; Represents the flow relationship of the production line; Representative workpiece Delivery time; Representative products The final process; This indicates the specified start time; This indicates the specified completion time; Indicates production line availability; Represents the amount of available resources; Representative process resource requirements; Inventory and batch parameters: Represents the target value for safety stock; Represents beginning inventory; Represents the planned shipment quantity; Represents economic batch size; Represents the threshold for small batches; This indicates whether it's a small batch; This indicates whether an order is being placed; This represents the sub-order relationship.

4. The intelligent scheduling method for automotive electronics manufacturing according to claim 3, characterized in that: The decision variables defined in step S100 include: Assignment and time series variables: Indicates the start of a process; Represents the start time of the process; Represents the time required to complete a process; Represents production line allocation; Indicates the sequence of processes; Resource and system variables: Indicates the use of markings on the production line; Represents the total time span; Represents capacity utilization rate; Inventory and batch variables: Represents inventory level; Represents production quantity; This indicates the presence of a batch; Represents batch size; This represents the switching cost.

5. The intelligent scheduling method for automotive electronics manufacturing according to claim 4, characterized in that: In step S200, the constraint conditions for priority 0 (highest priority) are as follows: Primary resource capacity constraints, expressed as ; Secondary resource capability constraints are represented as ; Process assignment constraints, represented as ; Production line exclusivity constraint, represented as ; Production line exclusivity constraint, represented as ; Process sequence constraints are represented as ; Production line flow constraints are represented as ; Processing time constraints are expressed as ; Sub-order sequence constraints are represented as .

6. The intelligent scheduling method for automotive electronics manufacturing according to claim 5, characterized in that: In step S200, the priority of other constraints is as follows: Priority 1: In production scheduling constraints, it is expressed as ; Priority 2: Specify date constraints, represented as ; Priority 3: Inventory level constraints are expressed as ; Initial inventory constraints, expressed as ; Minimum safety stock constraint, expressed as ; Economic quantity constraint, expressed as ; Mini-batch centralized constraints, denoted as ; Priority 4: Material kitting constraint, represented as ; Priority 5, 6: Capacity utilization constraints are expressed as capacity calculations: ； Priority 7: Switching the minimum constraint is represented as: ； Priority 8: SMT batch constraints, represented as ; Adjacent scheduling constraints are represented as ; Priority 9: Parallel process interval constraint, expressed as ; Priority 10: Order delivery constraints are represented as .

7. The intelligent scheduling method for automotive electronics manufacturing according to claim 6, characterized in that: The objective function in step S200 is expressed as follows: ; The weighting coefficient Adjustments are made based on constraint priority; for every 1-level difference in priority, the weights differ by a factor of 10.

8. The intelligent scheduling method for automotive electronics manufacturing according to claim 1, characterized in that: Step S300 specifically includes: S310: Initial population generation sub-step: An initial population is generated using a coarse-sorting strategy based on constraint priority. This strategy first satisfies the constraint with the highest priority. Orders are sorted and mapped to chromosome sequences according to the rules of prioritizing mass-produced products, prioritizing those with larger safety stock deviations, and prioritizing those with smaller absolute inventory quantities. The chromosomes are encoded in the order of order or product number. S320: Perform genetic iterative optimization on the population, in each generation: ① Use the roulette wheel selection operator to select parents based on individual fitness; ② Gene exchange is performed on the selected parent generation using a two-point crossover operator; ③ Apply a mutation operator to the crossover individuals, wherein the mutation is to randomly select a continuous gene segment in the chromosome and move it to the end of the sequence; ④ During the iteration process, for chromosome regions that do not meet specific constraints, targeted domain search based on parental gene information is used for adjustment; S330: Decoding and Evaluation Sub-steps: The optimized chromosome sequence is decoded into a specific production scheduling plan. The start time and completion time of each process are calculated based on the product process path and production line production rate. The fitness of the plan is evaluated using the objective function.

9. The intelligent scheduling method for automotive electronics manufacturing according to claim 8, characterized in that: The coarse-sorting strategy in step S310 is specifically as follows: ① Calculate the deviation between the current inventory of each mass-produced product and the minimum, standard, and maximum safety stock targets; ② Select products in ascending order of "first-level inventory deviation (difference from minimum safety stock)". If all deviations are positive, check the second-level and third-level inventory deviations in turn. ③ During the selection process, the availability of the target production line is monitored in real time; if it is unavailable, the current product is skipped. ④ If the inventory of all mass-produced products is sufficient, then new product or spare parts orders with a specified production date will be scheduled.

10. The intelligent scheduling method for automotive electronics manufacturing according to claim 9, characterized in that: The targeted domain search in step S320 specifically involves: Based on the location information of genes in the parent chromosome that do not meet the constraints, the order scheduling sequence at the corresponding locations is specifically adjusted to directly repair the constraint conflict.

Images