Wafer manufacturing production resource scheduling method based on gene expression programming
Through the method based on gene expression programming, a highly adaptable wafer manufacturing production resource scheduling strategy is constructed, which solves the problems of high computational complexity and poor adaptability in the existing technology, and realizes efficient and flexible production resource scheduling, reducing production costs.
Patent Information
- Application Number
- CN202510363170.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-26
- Publication Date
- 2025-06-06
AI Technical Summary
When the existing production scheduling methods deal with large-scale, high complexity and dynamically changing production environments, they have high computational complexity and poor adaptability, making it difficult to achieve efficient wafer manufacturing production resource scheduling in the semiconductor industry.
The wafer manufacturing production resource scheduling method based on gene expression programming is adopted. By obtaining the basic parameters of the historical production system, a production scheduling optimization model is constructed, and through encoding and decoding operations, a highly adaptable resource scheduling strategy is generated to achieve resource scheduling for the target wafer production tasks.
It effectively improves the data scheduling efficiency in the wafer manufacturing process, reduces production costs, and the generated scheduling rules are highly interpretable and scalable, and can flexibly respond to changes in the production environment and achieve dynamic and efficient task scheduling.
Smart Images
Figure CN120106510A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of production scheduling optimization, and more specifically to a wafer manufacturing production resource scheduling method based on gene expression programming. Background Art
[0002] In modern manufacturing, production scheduling is a key factor in determining production efficiency and resource utilization. Traditional production scheduling methods mainly rely on heuristic rules or mathematical optimization models, such as heuristic algorithms, integer programming, etc. However, these methods often face problems such as high computational complexity and poor adaptability when dealing with large-scale, complex, and dynamically changing production environments.
[0003] The core goal of production scheduling is to optimize the utilization of production resources, reduce production cycles, and improve overall production efficiency. In the actual production process, task scheduling faces many challenges, including changes in order priorities, uncertainty in equipment status, sudden failures in the production process, limited resources, and flexible requirements for production lines. In order to meet these challenges, intelligent optimization algorithms have become an important research direction in the field of production scheduling.
[0004] In recent years, intelligent optimization algorithms (such as gene expression programming (GEP), genetic algorithm (GA), ant colony optimization (ACO), reinforcement learning (RL), etc.) have gradually become important solutions to production scheduling problems. Gene expression programming (GEP) is an intelligent algorithm based on evolutionary computing. It can automatically learn and generate optimized scheduling rules and show strong adaptability in dynamic environments. GEP simulates the biological evolution process to generate more adaptable scheduling strategies, which can effectively solve complex and multi-objective optimization problems that are difficult to handle with traditional scheduling methods.
[0005] Existing production scheduling methods still have certain limitations, such as the fixed nature of scheduling rules, insufficient responsiveness to real-time data changes, and difficulty maintaining computing efficiency in large-scale production environments such as manufacturing, semiconductor production, and supply chains. Especially for wafer manufacturing in the semiconductor industry, enhancing the adaptability of scheduling solutions to changes in the production environment plays a key role in promoting the entire semiconductor industry to a higher level.
[0006] Therefore, how to improve data scheduling efficiency and reduce production costs in the wafer manufacturing process based on gene expression programming is an urgent problem that technical personnel in this field need to solve. Summary of the invention
[0007] In view of the above problems, the present invention provides a wafer manufacturing production resource scheduling method based on gene expression programming to at least solve some of the technical problems mentioned in the above background technology.
[0008] In order to achieve the above object, the present invention adopts the following technical solution:
[0009] The embodiment of the present invention provides a wafer manufacturing production resource scheduling method based on gene expression programming, comprising the following steps:
[0010] Obtain basic parameters of historical wafer manufacturing production systems;
[0011] Construct a production scheduling optimization model based on the gene expression programming algorithm;
[0012] The production scheduling optimization model is trained according to the basic parameters; the training process includes encoding operation and decoding operation;
[0013] Through the encoding operation, the basic parameters are converted into chromosomes in gene form, so that the gene expression programming algorithm can perform adaptive learning on the basic parameters;
[0014] Through the decoding operation, the encoded chromosome is converted into a resource scheduling strategy;
[0015] The trained production scheduling optimization model is used to realize resource scheduling for target wafer production tasks.
[0016] Further, the basic parameters of the historical wafer manufacturing production system include production task parameters and equipment status parameters;
[0017] The production task parameters include: task priority, processing time and resource requirements;
[0018] The device status parameters include: device availability, switching time and current load.
[0019] Furthermore, the obtaining of basic parameters of the historical wafer manufacturing production system includes:
[0020] Obtain task priority and processing time by reading configuration files of wafer manufacturing production system;
[0021] Obtain equipment availability and current load by monitoring equipment status;
[0022] Obtain equipment switching time through historical wafer manufacturing production data analysis or real-time sensor detection;
[0023] By analyzing the usage of wafer manufacturing production resources, resource requirements are estimated based on current task load and historical resource scheduling data.
[0024] Furthermore, during the encoding operation, a function set and a terminal set are constructed in combination with characteristics of wafer manufacturing resource scheduling problems.
[0025] Furthermore, the function set includes a plurality of symbols for performing operations.
[0026] Furthermore, the terminal set includes relevant parameters in the wafer manufacturing production resource scheduling process; the terminal set is represented as:
[0027]
[0028] ST ik =T setup,ik
[0029] CPT i =T proc,ij
[0030] WT i =S i -C i,j-1
[0031]
[0032] ET k =min{C i,j |i∈W k}
[0033] Where i is the number of the wafer; k is the number of the processing module; j is the number of the process; P i represents the set of all processes of wafer i; RT i represents the total processing time of all remaining steps of wafer i; T ij represents the processing time of the jth process corresponding to wafer i; ST ik represents the switching time required for processing module k to process wafer i; T setup,ik represents the machine preparation time required for processing module k to process wafer i; CPT i represents the processing time required for wafer i to be processed in the current process j; T proc,ij Indicates the preset time required for wafer i to process the current process j; WT i represents the waiting time of wafer i from the completion of the previous process to the start of the current process; S i Indicates the start time of the current process of wafer i; C i,j-1 MPT represents the completion time of the j-1th process of wafer i; k represents the average processing time of the process currently completed in processing module k; N k Indicates the number of processes that have been processed by processing module k; ET k represents the earliest available time of processing module k; C i,j represents the completion time of the jth process corresponding to wafer i; W k Represents the set of wafers that have been processed in processing module k.
[0034] Furthermore, during the decoding operation, the genes in the chromosome are interpreted by a gene expression programming algorithm to obtain a corresponding resource scheduling strategy.
[0035] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a wafer manufacturing production resource scheduling method based on gene expression programming, which has the following beneficial effects:
[0036] The wafer manufacturing production resource scheduling method based on gene expression programming disclosed in the present invention makes up for the problem of low operating efficiency of previous heuristic algorithms, and can effectively improve the data scheduling efficiency in the wafer manufacturing process and reduce production costs.
[0037] The wafer manufacturing production resource scheduling method based on gene expression programming disclosed in the present invention can generate scheduling rules that can explore the working principle of the heuristic algorithm, provide the interpretability of the relevant algorithm, and have strong scalability. In the face of emergency orders, mechanical failures and other situations, it can flexibly and quickly generate more ideal solutions, thereby realizing dynamic and efficient task scheduling in a complex production environment.
[0038] The invention discloses a wafer manufacturing production resource scheduling method based on gene expression programming, and proposes a hyper-heuristic wafer scheduling framework that can be automatically learned. The framework is flexible and convenient, applicable to various scenarios, and does not need to be re-run.
[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0041] Figure 1 A schematic flow chart of a method for wafer manufacturing production resource scheduling based on genetic expression programming provided in an embodiment of the present invention.
[0042] Figure 2 A schematic diagram of wafer manufacturing production resource scheduling provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0044] The embodiment of the present invention discloses a wafer manufacturing production resource scheduling method based on gene expression programming, such as Figure 1 As shown, the following steps are included:
[0045] S1. Obtain basic parameters of the historical wafer manufacturing production system;
[0046] S2. Construct a production scheduling optimization model based on the gene expression programming algorithm; train the production scheduling optimization model according to basic parameters; the training process includes encoding and decoding operations; through the encoding operation, the basic parameters are converted into chromosomes in the form of genes, so that the gene expression programming algorithm can adaptively learn the basic parameters; through the decoding operation, the encoded chromosomes are converted into resource scheduling strategies;
[0047] S3. Use the trained production scheduling optimization model to achieve resource scheduling for target wafer production tasks.
[0048] The wafer manufacturing production resource scheduling method based on gene expression programming provided by the embodiment of the present invention can maximize the utilization of production resources and optimize the task scheduling strategy to reduce switching costs and production waiting time. The characteristics of the present invention are: using the gene expression programming (GEP) algorithm to construct dynamic scheduling rules, and adaptively adjusting the scheduling plan according to the real-time changes of production tasks and equipment status to ensure the maximization of production efficiency and the optimality of resource allocation. It should be noted that this method can be used as a supplementary optimization means for existing production scheduling methods, and can be easily combined with existing optimization algorithms to provide a more comprehensive and efficient intelligent production scheduling optimization solution.
[0049] The above steps S1-S3 are only for the convenience of subsequent description and do not limit the specific execution order of each step. Next, each of the above steps will be described in detail.
[0050] In the above step S1, the basic parameters of the historical wafer manufacturing production system include production task parameters and equipment status parameters; wherein the production task parameters include: task priority, processing time and resource requirements; the equipment status parameters include: equipment availability, switching time and current load.
[0051] Obtain the basic parameters of the historical wafer manufacturing production system, specifically including: obtaining task parameters such as task priority and processing time by reading the configuration file of the wafer manufacturing production system; obtaining equipment availability and current load by monitoring the equipment status. This step can be monitored by the equipment status monitoring system; obtaining equipment switching time through historical wafer manufacturing production data analysis or real-time sensor detection; and estimating resource requirements based on current task load and historical resource scheduling data by analyzing the use of wafer manufacturing production resources.
[0052] In the above step S2, the production scheduling optimization model is trained according to the basic parameters; the training process includes encoding operation and decoding operation; wherein:
[0053] (1) Encoding operation:
[0054] Through coding operations, the basic parameters of wafer manufacturing production resource scheduling are converted into chromosomes in the form of genes. By converting the specific characteristics of the problem (such as processing time, switching time, waiting time, etc.) into the form of genes, the gene expression programming algorithm can adaptively learn the basic parameters. When designing the coding scheme, the function set (FS) and terminal set (TS) must be determined first:
[0055] Function Set (FS): Contains symbols for performing operations such as addition (+), subtraction (-), multiplication (*), division ( / ) and mathematical operations such as "exp". These operators can be used to combine and calculate various indicators in the production process, such as processing time, switching time, etc.
[0056] Terminal Set (TS): Contains parameters involved in the scheduling process, such as the remaining processing time of the task, the current load of the device, the waiting time, the switching time, etc. These are the specific input data of the problem and the basic elements of the gene expression. The terminal set designed in the scheduling rule training of the embodiment of the present invention is designed as follows:
[0057]
[0058] ST ik =T setup,ik
[0059] CPT i =T proc,ij
[0060] WT i =S i -C i,j-1
[0061]
[0062] ET k =min{C i,j |i∈W k}
[0063] Where i is the number of the wafer; k is the number of the processing module; j is the number of the process; P i represents the set of all processes of wafer i; RT i represents the total processing time of all remaining steps of wafer i; T ij represents the processing time of the jth process corresponding to wafer i; ST ik represents the switching time required for processing module k to process wafer i; T setup,ik represents the machine preparation time required for processing module k to process wafer i; CPT i represents the processing time required for wafer i to be processed in the current process j; T proc,ij Indicates the preset time required for wafer i to process the current process j; WT i represents the waiting time of wafer i from the completion of the previous process to the start of the current process; S i Indicates the start time of the current process of wafer i; C i,j-1 MPT represents the completion time of the j-1th process of wafer i; k represents the average processing time of the process currently completed in processing module k; N k Indicates the number of processes that have been processed by processing module k; ET k represents the earliest available time of processing module k; C i,j represents the completion time of the jth process corresponding to wafer i; W k Represents the set of wafers that have been processed in processing module k.
[0064] During the encoding process, each chromosome consists of multiple genes, each of which contains one or more functions and terminals, representing an operation in the scheduling process. One gene may represent "remaining processing time multiplied by switching time", while another gene may represent "the sum of task waiting time and device load". These genes will be combined to form a scheduling rule that can generate the optimal scheduling plan based on the current state of the production system.
[0065] (2) Decoding operation:
[0066] Decoding the best individual chromosome genes, the decoding process is to convert the encoded chromosomes into practical scheduling strategies. During the decoding process, gene expression programming (GEP) will calculate the corresponding scheduling rules by interpreting the genes in the chromosomes, and apply these rules to the actual scheduling problems. The decoded rules can be directly mapped to the actual operations in the production environment. One of the best degree rules obtained in the embodiment is trained as follows:
[0067]
[0068] Among them, RT represents the processing time of the remaining processes of the current wafer, which is the sum of all unfinished processes; ST represents the switching time required to select the processing module to process the wafer. The switching time may be different for different processing modules; CPT represents the processing time of the current process of the wafer, which is the time required for the current step; MPT represents the average processing time of the completed processes in the current processing module, which reflects the processing efficiency of the module and is obtained by calculating the number of processes processed in the module and the set of processed wafers; ET represents the earliest available time of the processing module, which is the minimum value of the final completion time of all wafers in the current module, and represents the next time point when the module can start processing.
[0069] In the above step S3, the basic parameters of the target wafer production system are input into the trained production scheduling optimization model. The production scheduling optimization model will determine the priority order of each production task based on information such as current production tasks, resource status and real-time changes; and perform task allocation and resource scheduling according to scheduling rules, and output the optimal strategy for resource scheduling of target wafer production.
[0070] In an embodiment of the present invention, the production scheduling optimization model can respond to changes in the production environment in real time. When special situations such as equipment failure, demand fluctuations or process sequence adjustments occur, it can automatically output the optimal production scheduling plan according to predetermined goals.
[0071] In actual operation, the order of production tasks, resource allocation scheme and equipment use arrangement can be dynamically adjusted according to the optimal strategy for resource scheduling of target wafer production, so as to ensure efficient and flexible production process. Through continuous iterative optimization. For details, please refer to Figure 2 As shown, after obtaining the wafer order demand, a wafer production plan is generated; the basic parameters corresponding to the wafer production plan are input into the trained production scheduling optimization model to generate the optimal scheduling strategy; based on the optimal scheduling strategy, resources are allocated to the production equipment, and the path of the material handling system is optimized to realize wafer processing.
[0072] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0073] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A wafer manufacturing production resource scheduling method based on gene expression programming, characterized in that: The steps include: Obtain basic parameters of historical wafer manufacturing production systems; Construct a production scheduling optimization model based on the gene expression programming algorithm; Training the production scheduling optimization model according to the basic parameters; During the training process, encoding operations and decoding operations are included; Through the encoding operation, the basic parameters are converted into chromosomes in gene form, so that the gene expression programming algorithm can perform adaptive learning on the basic parameters; Through the decoding operation, the encoded chromosome is converted into a resource scheduling strategy; The trained production scheduling optimization model is used to realize resource scheduling for target wafer production tasks.
2. According to claim 1, a wafer manufacturing production resource scheduling method based on gene expression programming is characterized in that: The basic parameters of the historical wafer manufacturing production system include production task parameters and equipment status parameters; The production task parameters include: task priority, processing time and resource requirements; The device status parameters include: device availability, switching time and current load.
3. A wafer manufacturing production resource scheduling method based on gene expression programming according to claim 2, characterized in that: The basic parameters of the historical wafer manufacturing production system are obtained, including: Obtain task priority and processing time by reading configuration files of wafer manufacturing production system; Obtain equipment availability and current load by monitoring equipment status; Obtain equipment switching time through historical wafer manufacturing production data analysis or real-time sensor detection; By analyzing the usage of wafer manufacturing production resources, resource requirements are estimated based on current task load and historical resource scheduling data.
4. The method for scheduling wafer manufacturing production resources based on genetic expression programming according to claim 1, characterized in that: During the encoding operation, a function set and a terminal set are constructed in combination with the characteristics of the wafer manufacturing resource scheduling problem.
5. The method for wafer manufacturing production resource scheduling based on gene expression programming according to claim 4, characterized in that: The function set contains a plurality of symbols that perform operations.
6. The method for wafer manufacturing production resource scheduling based on gene expression programming according to claim 4, characterized in that: The terminal set includes relevant parameters in the wafer manufacturing production resource scheduling process; the terminal set is represented as: ST ik =T setup,ik CPT i =T proc,ij WT i =S i -C i,j-1 AND k =min{C i,j |i∈W k } Where i is the number of the wafer; k is the number of the processing module; j is the number of the process; P i represents the set of all processes of wafer i; RT i represents the total processing time of all remaining steps of wafer i; T ij represents the processing time of the jth process corresponding to wafer i; ST ik represents the switching time required for processing module k to process wafer i; T setup,ik represents the machine preparation time required for processing module k to process wafer i; CPT i represents the processing time required for wafer i to be processed in the current process j; T proc,ij Indicates the preset time required for wafer i to process the current process j; WT i represents the waiting time of wafer i from the completion of the previous process to the start of the current process; S i Indicates the start time of the current process of wafer i; C i,j-1 MPT represents the completion time of the j-1th process of wafer i; k represents the average processing time of the process currently completed in processing module k; N k Indicates the number of processes that have been processed by processing module k; ET k represents the earliest available time of processing module k; C i,j represents the completion time of the jth process corresponding to wafer i; W k Represents the set of wafers that have been processed in processing module k.
7. The method for wafer manufacturing production resource scheduling based on gene expression programming according to claim 1, characterized in that: During the decoding operation, the genes in the chromosome are interpreted by a gene expression programming algorithm to obtain a corresponding resource scheduling strategy.