CN-121981483-A - Production scheduling data processing method and system based on template matching and process optimization

CN121981483ACN 121981483 ACN121981483 ACN 121981483ACN-121981483-A

Abstract

The invention relates to the technical field of industrial automation and production management, and provides a production scheduling data processing method and system based on template matching and process optimization, which are used for solving the problem of poor dynamic adaptability of the traditional production scheduling, wherein the method is characterized by acquiring a target row Cheng Moban and generating an initial production scheduling plan based on project characteristic data; in the execution of the initial production scheduling plan, a process coupling degree matrix and a resource load balancing factor are iteratively updated based on process execution progress data and resource state data, and accordingly, conflict simulation deduction is carried out on subsequent processes to judge whether to trigger the process local rearrangement, if so, a multi-constraint optimization model is built based on the process coupling degree matrix and the resource load balancing factor with the aim of minimizing scheduling disturbance, the affected process subset is locally and dynamically optimized, an adjusted process sequence is generated and issued for execution, and the self-adaption, the accuracy and the disturbance rejection capability of the production scheduling are remarkably improved.

Inventors

GAO ZUBO
Zhong Zuyong
Jiang Piyi
ZHANG ZHONGPAN
LI DASHUN
CHEN HENG

Assignees

广州市高波机电设备有限公司

Dates

Publication Date: 20260505
Application Date: 20260128

Claims (10)

1. The production schedule data processing method based on template matching and process optimization is characterized by comprising the following steps of: Responding to a scheduling instruction of a new project, and determining project characteristic data of the new project; acquiring a target row Cheng Moban based on the project characteristic data, generating an initial production scheduling plan based on the target row Cheng Moban, and issuing for execution after the initial production scheduling plan is confirmed; in the process of executing the initial production schedule, the following procedure optimization processes are iteratively executed: Acquiring process execution progress data and resource state data from a production terminal, updating a process coupling degree matrix for quantifying the dependency relationship between processes based on the process execution progress data and the resource state data, and updating a resource load balancing factor reflecting the real-time load state of resources; According to the updated process coupling degree matrix, the resource load balancing factor and the state of the current production scheduling plan, carrying out conflict simulation deduction on a process sequence to be executed subsequently, and judging whether to trigger process local rescheduling according to a deduction result; If the judgment is yes, taking the minimum scheduling disturbance as an optimization target, constructing a multi-constraint optimization model based on the process coupling degree matrix and the resource load balancing factor, and carrying out local dynamic optimization on the affected process subset to generate an adjusted process sequence; And after the adjusted sequence of working procedures is confirmed, the adjusted sequence of working procedures is issued to the corresponding production terminal so as to update the follow-up production execution plan.
2. The method of claim 1, wherein updating the process coupling matrix for quantifying inter-process dependencies comprises: Extracting a completion time stamp and a resource release identifier of the currently completed process from the process execution progress data; Calculating a dynamic offset of a theoretical earliest startable time of the affected subsequent process based on the completion timestamp; Updating the competition strength coefficient of the shared resource among corresponding procedures in the procedure coupling degree matrix according to the resource release identifier; and carrying out weighted fusion calculation on the dynamic offset and the competition strength coefficient by combining with preset logic dependent weights, and outputting an updated process coupling degree matrix.
3. The method for processing production schedule data based on template matching and process optimization of claim 2, wherein updating the resource load balancing factor reflecting the real-time load status of the resource comprises: Monitoring the current occupancy rate of each type of resource in a preset production resource set and the length of a task queue to be processed in real time; Calculating the predicted load saturation of each type of resource in a preset future time window based on the current occupancy rate and the length of the task queue to be processed; according to the predicted load saturation, calculating an initial load evaluation factor representing the overall resource tension degree by adopting a preset normalization function and a preset resource importance weighting coefficient; And comparing the initial load evaluation factor with the historical load trend data, and dynamically adjusting the initial load evaluation factor according to the comparison result to obtain a resource load balancing factor reflecting the real-time load state of the resource.
4. The method for processing production schedule data based on template matching and process optimization of claim 3, wherein comparing the initial load assessment factor with historical load trend data, and dynamically adjusting the initial load assessment factor according to the comparison result to obtain a resource load balancing factor reflecting a real-time load state of a resource, comprises: Calculating the deviation amount between the initial load evaluation factor and a load predicted value at the next moment predicted based on historical load trend data; Determining a corresponding dynamic adjustment coefficient according to a preset deviation interval to which the deviation value belongs, wherein the larger the deviation value is, the smaller the dynamic adjustment coefficient is; and carrying out weighted smoothing processing on the initial load evaluation factor based on the dynamic adjustment coefficient, wherein the calculation formula is as follows: L_final = α * L_new + (1 - α) * L_historical; wherein L_final is a resource load balancing factor after dynamic adjustment, L_new is the initial load evaluation factor, L_ historical is a historical load reference value determined according to historical trend data, alpha is the dynamic adjustment coefficient, and alpha is more than or equal to 0 and less than or equal to 1; and taking the calculated L_final as an updated resource load balancing factor.
5. The method for processing production schedule data based on template matching and process optimization of claim 1, wherein performing collision simulation deduction on a sequence of processes to be performed subsequently according to the updated process coupling degree matrix, the resource load balancing factor and the state of the current production schedule plan comprises: Acquiring the resource available capacity of each type of resource in the preset production resource set; Taking the current system time as a simulation starting point, taking a sequence of procedures to be executed subsequently, the procedure coupling degree matrix, the resource load balancing factor and the resource available capacity as inputs, and constructing a discrete event simulation model; In the discrete event simulation model, simulation is advanced according to time steps, logic and resource dependence among working procedures is analyzed according to the working procedure coupling degree matrix in each time step, and allocation and release of resources are scheduled according to the resource load balancing factors; In the simulation propulsion process, detecting and recording an event that the process starting time delay exceeds a first preset threshold value and an event that the resource load saturation exceeds a second preset threshold value as potential conflict points; And generating a conflict deduction report according to the number, the type and the severity of the potential conflict points.
6. The method for processing production schedule data based on template matching and process optimization of claim 5, wherein determining whether to trigger process local rescheduling based on the deduction result comprises: extracting the number of conflict points of N continuous time intervals in the future from the conflict deduction report to form a conflict point distribution sequence, wherein N is a natural number greater than 2; Inputting the conflict point distribution sequence into a preset time sequence prediction model, and predicting to obtain a future conflict development trend index; if the development trend index is greater than or equal to a preset trend deterioration threshold value, judging that the triggering procedure is locally rescheduled; If the development trend index is smaller than a preset trend deterioration threshold, analyzing the conflict deduction report, and extracting key performance indexes, wherein the key performance indexes comprise a predicted total construction period delay proportion, a key path conflict density and a bottleneck resource overload duration time; comparing the key performance indexes with corresponding preset trigger thresholds respectively; If the delay proportion of the estimated total construction period exceeds a preset first proportion threshold, or the collision density of the critical path exceeds a preset first density threshold, or the overload duration of the bottleneck resource exceeds a preset first time threshold, judging that the triggering procedure is locally rescheduled; Otherwise, determining not to trigger the process local rescheduling and continuing to execute the current production scheduling plan.
7. The method of claim 5 or 6, wherein constructing a multi-constraint optimization model based on the process coupling matrix and the resource load balancing factor comprises: Defining decision variables as adjusted planned starting time vectors for each process in the subset of affected processes; Quantifying the rank Cheng Raodong with a total adjustment to minimize the process plan start time as a first objective function; Taking the optimized lifting value of the maximized resource load balancing factor as a second objective function; taking a time logic constraint and a resource coupling constraint defined by a process coupling degree matrix as a first constraint condition; Taking the upper limit of the resource capacity defined by the resource available capacity and the resource load balancing factor as a second constraint condition; and combining the first objective function, the second objective function, the first constraint condition and the second constraint condition to construct a multi-objective mixed integer linear programming model and serve as a multi-constraint optimization model.
8. The method of claim 7, wherein the locally dynamically optimizing the subset of affected processes to generate the adjusted sequence of processes comprises: solving the multi-constraint optimization model by adopting a genetic algorithm with elite retention strategy, wherein the chromosome coding of the genetic algorithm adopts a real number coding mode of procedure starting time; Setting a crossover operator and a mutation operator of the genetic algorithm, wherein the crossover operator adopts ordered crossover based on process coupling degree, and the mutation operator adopts time window sliding mutation based on resource load; Generating a pareto optimal solution set through iterative computation of the genetic algorithm; And selecting a final solution from the pareto optimal solution set according to a preset decision preference function, and decoding a procedure starting time vector corresponding to the final solution into an adjusted procedure sequence.
9. The method for processing production schedule data based on template matching and process optimization of claim 1, wherein obtaining a target schedule template comprises: vectorizing the project characteristic data to obtain project characteristic vectors; Calculating multi-dimensional similarity between the project feature vector and the feature vector of each template in a preset project template library, wherein the multi-dimensional similarity comprises equipment configuration similarity, process complexity similarity and historical intersection expression similarity; according to a preset weighted fusion rule, fusing the multidimensional similarity into a comprehensive matching score; Selecting a template with the highest comprehensive matching score as a first candidate template, and recording the highest score corresponding to the template; judging whether the highest score is larger than or equal to a preset score threshold value; if the highest score is greater than or equal to the preset score threshold, determining the first candidate template as a target row Cheng Moban; If the highest score is smaller than the preset score threshold, combining the project characteristic data with the historical project characteristic data of the historical executed project to form a characteristic data set to be analyzed; Performing cluster analysis on the feature data set to be analyzed by adopting a preset clustering algorithm to obtain a project cluster division result containing the new project; Identifying a cluster to which the newly built item belongs, and fusing and summarizing historical production scheduling plans of all the historical executed items in the cluster to generate an initial row Cheng Moban; After the initial rank Cheng Moban is validated, it is determined as the target scheduling template.
10. A production scheduling data processing system based on template matching and process optimization, comprising: the first determining module is used for responding to the scheduling instruction of the new project and determining project characteristic data of the new project; The first generation module is configured to obtain a target rank Cheng Moban based on the project feature data, generate an initial production scheduling plan based on the target rank Cheng Moban, and issue the initial production scheduling plan for execution after the initial production scheduling plan is confirmed; a first iteration module for iteratively performing a process optimization process during execution of the initial production schedule based on the following sub-modules, and comprising: the first updating sub-module is used for acquiring process execution progress data and resource state data from the production terminal, updating a process coupling degree matrix for quantifying the dependency relationship between the processes based on the process execution progress data and the resource state data, and updating a resource load balancing factor reflecting the real-time load state of the resource; the first deduction submodule is used for carrying out conflict simulation deduction on a sequence of a procedure to be executed subsequently according to the updated procedure coupling degree matrix, the resource load balancing factor and the state of the current production scheduling plan, and judging whether to trigger the procedure to be locally rescheduled according to a deduction result; The first optimization sub-module is used for constructing a multi-constraint optimization model based on the process coupling degree matrix and the resource load balancing factor by taking the minimized scheduling disturbance as an optimization target if the judgment is yes, carrying out local dynamic optimization on the affected process subset, and generating an adjusted process sequence; And the first issuing submodule is used for issuing the regulated sequence to the corresponding production terminal after confirming the regulated sequence so as to update the subsequent production execution plan.

Description

Production scheduling data processing method and system based on template matching and process optimization Technical Field The invention relates to the technical field of industrial automation and production management, in particular to an intelligent manufacturing scheduling technology based on data driving and model optimization, and specifically relates to a production scheduling data processing method and system based on template matching and process optimization. Background ‌ ‌ Production scheduling is the core of manufacturing enterprises, particularly electronic equipment (cabinet) and other enterprises designed according to orders (ETO), and aims to convert contract orders into executable production instructions and overall arrange procedures, resources and time so as to ensure on-schedule, quality guarantee and low-cost delivery of projects. With the development of intelligent manufacturing, production systems have placed higher demands on the dynamic adaptability, accuracy and immunity of the schedules. In the conventional technology, the production schedule mainly adopts static schedule based on a fixed template or adjustment based on manual experience. Based on the method of manual experience, a dispatcher receives anomalies through an instant communication tool (such as WeChat), then manually adjusts the plan by experience, the response is delayed, and global optimization is difficult to ensure. The two methods have the common limitation that the automatic quantitative sensing of the dynamic coupling relation between the working procedures and the resource competition state is not established, and the advanced automatic adjustment based on prediction and prevention is not formed, so that the scheduling is passive and lagged. In summary, the fundamental problem of the conventional scheduling is that the production scheduling of static or manual intervention lacks dynamic quantitative modeling and prospective optimizing capability for process coupling and resource constraint, so that the scheduling plan has poor dynamic adaptability in the process of production execution, and cannot realize adaptive early adjustment with minimum disturbance. Therefore, how to solve the problems of poor dynamic adaptability and large adjustment disturbance of the scheduling plan caused by the fact that the coupling relation of the process and the resource load state cannot be quantized in real time and the prospective optimization is performed in the conventional production scheduling is a technical problem to be solved. Disclosure of Invention Aiming at the technical problems, the invention provides a production scheduling data processing method and a system based on template matching and process optimization, which can remarkably improve the dynamic self-adaption and the accuracy of scheduling in a production environment. The invention provides a production scheduling data processing method based on template matching and process optimization, which comprises the following steps of responding to scheduling instructions of new projects, determining project characteristic data of the new projects, acquiring a target schedule Cheng Moban based on the project characteristic data, generating an initial production scheduling plan based on the target schedule Cheng Moban, issuing for execution after the initial production scheduling plan is confirmed, iteratively executing a process optimization process in the process of executing the initial production scheduling plan, acquiring process execution progress data and resource state data from a production terminal, updating a process coupling degree matrix for quantifying a process dependency relationship based on the process execution progress data and the resource state data, updating a resource load balancing factor reflecting a resource real-time load state, carrying out conflict simulation deduction on a sequence to be executed subsequently according to the updated process coupling degree matrix, the resource load balancing factor and the state of the current production scheduling plan, judging whether a process is locally rescheduled according to a deduction result, and carrying out a process optimization sequence based on the fact that the process is the minimum, carrying out the process is the process optimization, constructing a dynamic optimization model based on the target schedule and the optimized sequence, and carrying out dynamic adjustment to the process optimization model, and setting the process optimization is carried out to the corresponding process optimization model based on the minimum disturbance. The invention also provides a production scheduling data processing system based on template matching and process optimization, which comprises a first determining module, a second determining module and a processing module, wherein the first determining module is used for responding to a scheduling instruction of a new project and determining project cha