Mold parameter adjusting method and system based on multi-objective optimization algorithm

Abstract

The invention discloses a mold parameter adjustment method and system based on a multi-objective optimization algorithm, and relates to the technical field of mold manufacturing. The method comprises the following steps: calling a production database, carrying out production state division, and establishing a production state division result; executing process data acquisition, and synchronizing a process data acquisition result to a data processing center through an MQTT protocol; matching a process data acquisition result with a production state division result by using a data processing center, and constructing a real-time dynamic weight; updating the multi-objective evaluation function by using the real-time dynamic weight, and establishing a configuration optimization priority; adjusting and optimizing the mold parameters according to the multi-objective evaluation function and the optimizing priority, and optimizing the mold parameters based on an adjusting and optimizing result. The technical problems that in the prior art, the die production efficiency is low, and the quality is unstable are solved, the die parameters are adjusted through the multi-objective optimization algorithm and the real-time dynamic weight, and the technical effect of improving the die production efficiency and the manufacturing quality is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of mold manufacturing, and in particular to a mold parameter adjustment method and system based on a multi-objective optimization algorithm. Background Art

[0002] In the modern mold manufacturing industry, with the increase in product complexity and the intensification of market competition, higher requirements are placed on the refined management and efficiency of the mold manufacturing process. As a key tool in industrial production, the rationality of the mold parameter setting directly affects the quality and production efficiency of the product. However, traditional mold parameter adjustment methods often rely on experience judgment or fixed patterns. The debugging process is time-consuming and it is difficult to take into account both production efficiency and manufacturing quality at the same time, thus failing to adapt to the rapidly changing production environment and diversified production needs. In the actual production process, different production states and production goals have different requirements for mold parameters, and it is difficult to achieve the ideal effect through a single optimization method. Therefore, there is an urgent need for a mold parameter adjustment method based on a multi-objective optimization algorithm to improve production efficiency and product quality and meet complex production needs. Summary of the invention

[0003] The present application provides a mold parameter adjustment method and system based on a multi-objective optimization algorithm, which solves the technical problems of low mold production efficiency and unstable quality in the prior art.

[0004] In view of the above problems, the present application provides a mold parameter adjustment method and system based on a multi-objective optimization algorithm.

[0005] In a first aspect of the present application, a mold parameter adjustment method based on a multi-objective optimization algorithm is provided, the method comprising: Access the management equipment of the factory, call the production database, and divide the production status based on the production database to establish the production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight ratios reflecting the production target; use the sensor network to perform process data collection, and synchronize the process data collection results to the data processing center through the MQTT protocol, and the data processing center stores the production status division results; use the data processing center to match the process data collection results with the production status division results, and construct real-time dynamic weights with the static weight group corresponding to the matching results; use the real-time dynamic weights to update the multi-objective evaluation function, and establish the configuration optimization priority; adjust and optimize the mold parameters according to the multi-objective evaluation function and the optimization priority, and optimize the mold parameters based on the adjustment optimization result.

[0006] The second aspect of the present application provides a mold parameter adjustment system based on a multi-objective optimization algorithm, the system comprising: A production status division module, the production status division module is used to access the factory's management equipment, call the production database, and perform production status division based on the production database to establish a production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target; a data acquisition module, the data acquisition module is used to perform process data acquisition using a sensor network, and synchronize the process data acquisition results to a data processing center through the MQTT protocol, and the data processing center stores the production status division results; a matching module, the matching module is used to match the process data acquisition results with the production status division results using the data processing center, and construct real-time dynamic weights with the static weight group corresponding to the matching results; an update module, the update module is used to update the multi-objective evaluation function using real-time dynamic weights, and establish a configuration optimization priority; an optimization module, the optimization module is used to adjust and optimize the mold parameters according to the multi-objective evaluation function and the optimization priority, and optimize the mold parameters based on the adjusted optimization results.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, access the management equipment of the factory, call the production database, and divide the production status based on the production database to establish the production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target. Then, the sensor network is used to perform process data collection, and the process data collection results are synchronized to the data processing center through the MQTT protocol, and the data processing center stores the production status division results. Further, the data processing center matches the process data collection results with the production status division results, and constructs real-time dynamic weights with the static weight group corresponding to the matching results. Then, the multi-objective evaluation function is updated using the real-time dynamic weight, and the configuration optimization priority is established. Finally, the mold parameters are adjusted and optimized according to the multi-objective evaluation function and the optimization priority, and the mold parameters are optimized based on the adjustment optimization results. The technical problems of low mold production efficiency and unstable quality in the prior art are solved, and the mold parameters are adjusted by multi-objective optimization algorithm and real-time dynamic weights, so as to achieve the technical effect of improving mold production efficiency and manufacturing quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1A schematic diagram of a process flow of a mold parameter adjustment method based on a multi-objective optimization algorithm provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a mold parameter adjustment system based on a multi-objective optimization algorithm provided in an embodiment of the present application.

[0010] Explanation of the accompanying drawings: production status division module 11, data collection module 12, matching module 13, updating module 14, optimization module 15. DETAILED DESCRIPTION

[0011] The present application solves the technical problems of low mold production efficiency and unstable quality in the prior art by providing a mold parameter adjustment method and system based on a multi-objective optimization algorithm.

[0012] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0013] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.

[0014] Embodiment 1, as Figure 1 As shown, the present application provides a mold parameter adjustment method based on a multi-objective optimization algorithm, wherein the method includes: Access the management equipment of the factory, call the production database, and divide the production status based on the production database to establish the production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target.

[0015] By accessing the factory's management equipment, the production database can be retrieved to obtain production-related data, including but not limited to production plans, production progress, equipment status, product quality feedback, material supply status, etc.

[0016] In the mold manufacturing process, different production stages (such as pre-processing, finishing, debugging, etc.) have different requirements for production efficiency, product quality, and production cost. By analyzing the production data, the production process can be divided into multiple different states, each of which has its own specific production goals and requirements. For example, in the initial stage, more attention may be paid to speed, while in the later finishing stage, more attention may be paid to quality. Each production state division result corresponds to a static weight group, which reflects the weight proportion of each production goal in the current production state. For example, the static weight group in a certain state may include product quality, mold life, production efficiency, production cost, etc. The setting of static weight groups is usually based on historical data, expert experience, and corporate strategic goals to ensure that production parameters can be optimized in a targeted manner under different production states. Each static weight group is a set of weight proportions, in which each weight corresponds to a specific production goal. By adjusting these weights, enterprises can achieve flexible balance and optimization of production goals under different production states.

[0017] The sensor network is used to perform process data collection, and the process data collection results are synchronized to the data processing center through the MQTT protocol. The data processing center stores the production status division results.

[0018] The sensor network is composed of sensors distributed at key nodes of the production line. These sensors are responsible for real-time monitoring of various process parameters in the production process, including temperature, pressure, vibration, speed, humidity, position, etc. MQTT is a lightweight message transmission protocol suitable for data transmission in environments with limited bandwidth and unreliable networks. Through the MQTT protocol, the data collected by the sensors can be transmitted to the remote data processing center in real time to ensure efficient synchronization of data and meet the real-time requirements of mold parameter adjustment. The data processing center not only stores the data collected in real time, but also saves the results of production status division, so that it can combine real-time data to make intelligent matching and dynamic optimization of production status.

[0019] The data processing center is used to match the process data collection results with the production status division results, and the real-time dynamic weights are constructed with the static weight groups corresponding to the matching results.

[0020] The data processing center matches the collected process data with the production state division results. Each production state division will have a set of preset process parameter ranges (such as temperature, pressure, etc.). If the real-time collected process data falls within the parameter range of a certain state, the current production state is considered to belong to the division result. Each production state division result corresponds to a static weight group, which reflects the relative importance of different production goals (such as efficiency, quality, cost, etc.) in this state. By matching the process data with the production state division results, it is possible to determine which division result the current production state belongs to, and then call the static weight group of that state.

[0021] Since various process parameters in the actual production process will fluctuate over time, and the priority of production goals may be affected by real-time changes, static weights alone cannot adapt to the rapidly changing needs of the production process. Therefore, based on the matching results, it is necessary to convert the static weight group into real-time dynamic weights.

[0022] Furthermore, the construction of real-time dynamic weights using static weights corresponding to matching results also includes: Perform a time series change analysis on the process data collection results, perform a time series prediction of the production status based on the time series change analysis results, and establish a time series prediction result; perform data segmentation based on the process data collection results and the time series prediction results, and generate a backtracking compensation weight and a predictive compensation weight for the target corresponding to the static weight group based on the data segmentation result; perform corresponding static weight group compensation based on the backtracking compensation weight and the predictive compensation weight, and construct a real-time dynamic weight with the compensation result.

[0023] Specifically, in the production process, process data (such as temperature, pressure, etc.) will change over time. By performing time series change analysis on the process data collection results, the changing trend of process parameters can be identified; based on the time series change analysis results, production status time series prediction is performed to establish time series prediction results, that is, to predict changes in production status in the future. For example, the analysis results may predict that the temperature will exceed the preset range or the pressure will fluctuate at a certain time in the future; based on the process data and time series prediction results, the data is segmented to distinguish historical data, real-time data and future predicted data. According to these data in different time dimensions, two types of complementary Compensation weights: retrospective compensation weights and predictive compensation weights; retrospective compensation weights are based on historical process data, analyzing deviations or deficiencies in previous production processes, and retrospectively compensating for corresponding targets. For example, if it is found that production quality does not meet standards for a period of time, the quality weight can be increased in subsequent weight adjustments to compensate for previous deficiencies; predictive compensation weights are based on time series forecasting results, and the weights of production targets are adjusted in advance for possible production changes in the future. For example, if it is predicted that the temperature will fluctuate in the future, the production quality weight may be increased in advance and the production efficiency weight may be reduced to ensure that quality will not be affected when fluctuations occur.

[0024] According to the retrospective compensation weight and the forecast compensation weight, the static weight group is compensated and adjusted. Specifically, if some goals (such as quality and efficiency) in the historical production status do not meet expectations, retrospective compensation is performed by increasing the weight of the corresponding goals. For example, if the product quality is poor for a period of time during the production process, the system will temporarily increase the quality-related weights to make up for the previous deficiencies; foresee future changes based on the time series prediction results and adjust the weights in advance. For example, if it is predicted that future production efficiency may decrease due to equipment failure, the system will compensate for the efficiency target in advance to ensure that other goals can be maintained as stable as possible when the efficiency decreases. Real-time dynamic weights are constructed based on retrospective compensation weights and forecast compensation weights. They have higher flexibility and adaptability, and can adjust mold parameters in real time during the production process to make production more stable and efficient.

[0025] The multi-objective evaluation function is updated using real-time dynamic weights, and the configuration optimization priority is established.

[0026] The multi-objective evaluation function is a mathematical model that comprehensively evaluates the key objectives in the production process (such as production efficiency, product quality, production cost, etc.). Its purpose is to find a balance between different production objectives to ensure the best overall optimization effect. The multi-objective evaluation function can generally be expressed as: , where F is the comprehensive result of the multi-objective evaluation function, , , …, is the evaluation value of each production target (such as efficiency, quality, cost, etc.), , , …, The weights of each target are adjusted based on the real-time dynamic weights. When the real-time dynamic weights are constructed, the system will introduce them into the multi-objective evaluation function and update the weight ratios of each production target. The introduction of real-time dynamic weights can ensure the flexibility of the multi-objective evaluation function in the production process, enabling it to quickly respond to changes in the production environment and achieve dynamic optimization. After updating the multi-objective evaluation function, it is necessary to establish optimization priorities based on the importance of different production targets. The optimization priorities determine which targets should be given priority and which targets can be appropriately compromised when adjusting mold parameters.

[0027] Furthermore, the method of updating the multi-objective evaluation function using real-time dynamic weights and establishing configuration optimization priorities also includes: Establish a basic objective function, which includes a production efficiency function, a mold life function, a material utilization function, and a product quality function; weighted fit the basic objective function through the real-time dynamic weight to establish a multi-objective evaluation function; perform a weight change rate analysis on the real-time dynamic weight, and establish an optimization priority based on the change value of the weight change rate.

[0028] In the multi-objective optimization process, different production objectives often have different evaluation criteria and calculation methods, and multiple basic objective functions are established to facilitate weighting and optimization in the subsequent process. The basic objective functions include production efficiency function, mold life function, material utilization function, and product quality function; the production efficiency function is used to evaluate the production volume or production speed per unit time, and production efficiency = output / unit time; the mold life function is used to evaluate the service life of the mold, and mold life = mold usage time / total life; the material utilization function is used to measure the efficiency of material utilization in the production process, and material utilization = effective material / total material consumption; the product quality function is used to evaluate the quality of the product, and product quality = number of qualified products / total number of products. These basic objective functions are weighted and fitted through real-time dynamic weights to form a multi-objective evaluation function. ,in, , , , For real-time dynamic weight, is the corresponding basic objective function.

[0029] The weight change rate analysis is performed on the real-time dynamic weights. The purpose of the weight change rate analysis is to observe the changing trend of the real-time dynamic weights at different time points, so as to identify which targets have the largest weight change and which targets have the smallest weight change. The calculation formula of the weight change rate can be expressed as , is the weight value at a certain time point, is the weight value at the next time point, is the rate of change of the weight of the target. According to the size of the weight change rate, the priority of each target is sorted. For example, the target with the largest weight change rate (such as product quality) will be optimized first, while the target with a smaller weight change (such as material utilization) can be optimized later.

[0030] The mold parameters are adjusted and optimized according to the multi-objective evaluation function and the optimization priority, and the mold parameters are optimized based on the adjustment optimization result.

[0031] After deriving the multi-objective evaluation function and optimization priority, the mold parameters are adjusted and optimized. Specifically, according to the optimization priority, the current target to be optimized is determined, and an optimization algorithm such as a genetic algorithm or a particle swarm algorithm is selected to adjust the current mold parameters so that they gradually tend to the optimal solution. In each optimization iteration, the mold parameters are fine-tuned according to the multi-objective evaluation function and the current optimization priority. The mold parameters are continuously adjusted and optimized to ensure the balance and improvement of various production goals. The final parameter setting is based on the optimization results to ensure that the mold parameters reach the optimal state in actual production, which not only improves production efficiency, but also ensures product quality and the long-term service life of the mold.

[0032] Furthermore, the adjusting and optimizing of mold parameters according to the multi-objective evaluation function and optimization priority also includes: An initial solution space is established, wherein the initial solution space contains initial solutions of a first proportion and initial solutions of a second proportion, wherein the initial solutions of the first proportion are solution sets constructed by historical parameters, and the initial solutions of the second proportion are random solution sets, and the first proportion is greater than the second proportion, and the sum of the first proportion and the second proportion is 1, and each solution in the initial solution space represents a parameter combination of multiple objectives; the fitness of the solutions in the initial solution space is calculated using the multi-objective evaluation function, and a leader solution set is established according to the fitness calculation result, wherein the leader solution set includes at least one historical parameter solution and one random solution; the remaining solutions in the initial solution space are searched and updated based on the leader solution and the optimization priority, and the leader solution set is updated, and the initial solution space is reconstructed with the updated leader solution set and the search update result; a search iteration is performed, and the adjustment optimization is completed according to the search iteration result.

[0033] Before optimizing mold parameters, it is necessary to construct an initial solution space, which contains multiple candidate solutions, each of which represents a possible combination of mold parameters. The first proportion of initial solutions is a solution set constructed based on historical parameters. This part of the solution set comes from the optimization results or empirical data in the system's historical production process. Since historical data has certain experience and reliability, these solutions are usually closer to the optimal solution in the actual production process. The first proportion of initial solutions accounts for the majority, ensuring that the system can draw on previous experience in the initial stage and quickly approach the optimal solution. The second proportion of initial solutions is a randomly generated solution set. In order to avoid falling into the local optimum, the system generates a part of the initial solutions through a random algorithm. These random solutions can provide more diversity for the optimization process, thereby increasing the coverage of the solution space. Among them, the first proportion is greater than the second proportion, and the sum of the two is 1, that is, the first proportion + the second proportion = 1, so as to ensure that the solution space is mainly based on historical parameters, while retaining a certain degree of randomness, enhancing the global search capability of the system.

[0034] The multi-objective evaluation function is used to calculate the fitness of each solution in the solution space. The purpose of the fitness calculation is to evaluate the performance of each solution in meeting the multi-objective optimization requirements, that is, to determine its comprehensive score on various objectives (such as production efficiency, mold life, material utilization, product quality, etc.) through the evaluation function. The higher the fitness score, the better the solution performs in multi-objective optimization. After the fitness calculation is completed, a leader solution set is established based on the calculation results. The solution set contains solutions with higher fitness. In order to ensure the diversity of the search, the leader solution set contains at least one historical parameter solution and one random solution. The historical parameter solution represents the optimization solution derived from historical data, which usually has higher stability and reliability. The fitness of the random solution may be lower, but it helps to expand the search space and avoid falling into the local optimum.

[0035] After the leader solution set is established, the remaining solutions in the initial solution space are searched and updated. The solutions in the leader solution set are used to guide the update of the remaining solutions, that is, the search and update of the solutions will be gradually optimized based on the leader solution; through the optimization priority, determine which goal needs to be optimized first, adjust the direction of the search and update according to the priority, and ensure that the most important goal is met first during the search process. For example, when the priority of product quality is high, the solutions that can improve the quality-related parameters will be selected for update. The updated solution will continue to be calculated for fitness through the multi-objective evaluation function, and the leader solution set will be continuously updated to make the entire search process tend to the optimal. After a search update based on the leader solution, the system will reconstruct the initial solution space based on the updated leader solution set and the search update results; the reconstructed solution space contains the updated solution and the optimized leader solution set, ensuring that in each iteration, the solution space can continuously approach the global optimal solution. The entire adjustment optimization process is achieved through multiple search iterations. After each iteration, the solution space will be readjusted based on the leader solution and the optimization priority, and the optimization search will continue. After multiple iterations of searching and optimization, the system selects the optimal solution in the initial solution space and completes the optimization of the mold parameters.

[0036] Furthermore, the performing of the search iteration further includes: An iterative observation layer is established, and the iterative observation layer is used to observe and evaluate the results of each round of search iterations; when the observation and evaluation results meet a preset trigger threshold, a leader solution set constraint instruction is generated, and the preset trigger threshold is a threshold value where the number of iterations is within a preset number range and the search variability is lower than a preset variability value; the leader solution set generation constraint for the next round of iteration is performed according to the leader solution set constraint instruction.

[0037] An iterative observation layer is established to observe and evaluate the iteration results after each round of search iteration. The iterative observation layer evaluates the effectiveness and variability of the current iteration by analyzing the solution update, fitness change, search range and other key indicators in each round of iteration, including: fitness improvement evaluation (monitoring whether the fitness value of the solution in each round of iteration is significantly improved), search range evaluation (evaluating the diversity of the current search to avoid all solutions being too concentrated in a certain solution space, resulting in falling into a local optimum).

[0038] In order to ensure that the optimization process is always effective, the system sets a preset trigger threshold for the iterative observation layer; when the observation evaluation results reach or meet this trigger threshold, corresponding constraint measures will be taken. The preset trigger threshold is the threshold value when the number of iterations is within the preset number range and the search variability is lower than the preset variability value, where the search variability refers to the solution variability in each round of iteration, that is, the magnitude of change in the solution during the iteration process. If the solution changes too little during the search process, it means that the search space lacks diversity, which may cause the optimization process to be too concentrated and unable to break through the local optimum. The preset variability value is a threshold defined by the system. When the search variability is lower than this threshold, it is considered that the search is stagnant or local optimal and needs to be adjusted.

[0039] When the iteration observation layer determines that the result of a certain round of iterations has reached the preset trigger threshold, it will generate a leader solution set constraint instruction, which is used to constrain the generation of the leader solution set in the next round of iterations to improve search efficiency and optimization effect. In the next round of iterations, the leader solution update is restricted, such as limiting the frequent update of the leader solution, optimizing the search path, adjusting the proportion of random solutions, etc., to ensure that the optimization direction is more focused on the possible optimal solution space. Through this constraint, unnecessary solution updates can be reduced and resource waste can be avoided.

[0040] Furthermore, the method further comprises: Perform production monitoring and establish a real-time response database; perform response evaluation on the real-time response database and establish a reinforcement learning controller; and use the reinforcement learning controller to update and iterate a multi-objective evaluation function.

[0041] During the production process, the system performs real-time production monitoring, continuously monitors various process parameters, equipment status, production efficiency, product quality, etc., and dynamically stores the monitored data in the real-time response database. These monitoring data provide real-time feedback for the optimization process to ensure that the system can respond to changes in production in a timely manner. Based on the data in the real-time response database, the response situation in the production process is evaluated. By analyzing historical responses and real-time feedback, the fit between the current optimization effect and the production status is judged; the response evaluation includes process response time and production target response. The process response time refers to whether the system makes rapid and effective adjustments to process parameters when monitoring changes in the production process. The production target response refers to the evaluation of the real-time response effect of the system in multi-objective optimization, such as whether production efficiency, mold life, material utilization and product quality have been effectively improved. Based on the real-time data and response evaluation of production monitoring, a reinforcement learning controller is constructed. The reinforcement learning controller continuously learns the optimization strategy in the production process, analyzes the response effect of the system to process changes, and gradually improves the optimization strategy.

[0042] The reinforcement learning controller will continuously adjust the multi-objective evaluation function dynamically and update iteratively based on the real-time data and process feedback from production monitoring. By learning the actual data and response results in production, the reinforcement learning controller can dynamically adjust the weights in the multi-objective evaluation function. For example, at a certain stage, production efficiency may be the most critical goal, and the controller will automatically increase the weight of efficiency-related goals; when the mold life is detected to be excessively worn, the controller will increase the weight of life-related goals. The reinforcement learning controller can not only dynamically adjust the weights, but also update the multi-objective evaluation function according to the feedback from production in each iteration. It constantly conducts strategy trials in the production process and evaluates the pros and cons of the current evaluation function based on the fitness of the feedback; through iterative updates, the system can gradually converge to the optimal multi-objective evaluation function to ensure the balance between different production goals. With the progress of multiple rounds of iterations, the reinforcement learning controller gradually optimizes its own strategy and applies it to the adjustment and optimization of mold parameters. In each optimization process, the controller will refer to the feedback from the previous round and continuously improve to ensure that the next round of adjustments is more accurate and efficient. After multiple iterative updates of the reinforcement learning controller, the system will eventually converge to an optimal multi-objective evaluation function and a set of optimal mold parameter combinations.

[0043] Furthermore, the method further comprises: Establish a multi-level warning mapping set; use the real-time response database to perform warning trigger analysis on the multi-level warning mapping set; and issue a warning alert based on the mapping warning level corresponding to the warning trigger analysis result.

[0044] The multi-level warning mapping set is a warning system for defining different degrees of production anomalies or potential problems; by setting thresholds and classifying different process parameters (such as temperature, pressure, production efficiency, mold life, etc.), the warning is divided into different levels to ensure that the system can detect and issue an alarm in time before the anomaly occurs. According to the severity of the anomaly, the warning is divided into different levels. For example, the first-level warning: minor anomalies, such as a slight decrease in production efficiency, mold temperature approaching the upper limit, etc., may not immediately affect production, but attention should be paid; the second-level warning: moderate anomalies, such as reduced material utilization, equipment beginning to show signs of wear, and production parameters need to be adjusted in time to avoid further deterioration; the third-level warning: severe anomalies, such as mold life approaching the limit or product quality significantly reduced, at which time the system needs to intervene quickly and may need to suspend production for maintenance. Each warning level is mapped to a specific production parameter threshold. For example, if the temperature exceeds a certain critical value, a second-level warning is triggered; if the production efficiency drops by more than a certain percentage, a first-level warning is triggered.

[0045] Using the data in the real-time response database, the current real-time process parameters (such as temperature, pressure, efficiency, etc.) are compared with the thresholds in the multi-level warning mapping set. For example, when the mold temperature approaches the set high temperature threshold, the system will determine whether to trigger an early warning based on the temperature change trend; when the production efficiency is lower than the preset minimum efficiency value, the system will evaluate the production status and make corresponding early warning judgments. When the early warning trigger analysis detects an abnormality, the system will issue an early warning according to the severity of the abnormality and the corresponding mapping early warning level. The early warning output methods of different levels will also be different to ensure that the system can respond quickly and take appropriate measures. Through the multi-level early warning mechanism, the system can identify potential problems in advance and automatically adjust the optimization strategy to prevent the abnormality from further affecting production efficiency, product quality and mold life.

[0046] In summary, the embodiments of the present application have at least the following technical effects: First, access the management equipment of the factory, call the production database, and divide the production status based on the production database to establish the production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target. Then, the sensor network is used to perform process data collection, and the process data collection results are synchronized to the data processing center through the MQTT protocol, and the data processing center stores the production status division results. Further, the data processing center matches the process data collection results with the production status division results, and constructs real-time dynamic weights with the static weight group corresponding to the matching results. Then, the multi-objective evaluation function is updated using the real-time dynamic weight, and the configuration optimization priority is established. Finally, the mold parameters are adjusted and optimized according to the multi-objective evaluation function and the optimization priority, and the mold parameters are optimized based on the adjustment optimization results. The technical problems of low mold production efficiency and unstable quality in the prior art are solved, and the mold parameters are adjusted by multi-objective optimization algorithm and real-time dynamic weights, so as to achieve the technical effect of improving mold production efficiency and manufacturing quality.

[0047] Embodiment 2 is based on the same inventive concept as the mold parameter adjustment method based on the multi-objective optimization algorithm in the previous embodiment. Figure 2 As shown, the present application provides a mold parameter adjustment system based on a multi-objective optimization algorithm, wherein the system includes: A production status division module 11, the production status division module 11 is used to access the factory's management equipment, call the production database, and perform production status division based on the production database to establish a production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target; a data acquisition module 12, the data acquisition module 12 is used to perform process data acquisition using a sensor network, and synchronize the process data acquisition results to a data processing center through the MQTT protocol, and the data processing center stores the production status division results; a matching module 13, the matching module 13 is used to match the process data acquisition results with the production status division results using the data processing center, and construct a real-time dynamic weight with the static weight group corresponding to the matching result; an update module 14, the update module 14 is used to update the multi-objective evaluation function using the real-time dynamic weight, and establish a configuration optimization priority; an optimization module 15, the optimization module 15 is used to adjust and optimize the mold parameters according to the multi-objective evaluation function and the optimization priority, and optimize the mold parameters based on the adjustment optimization result.

[0048] Furthermore, the matching module 13 is used to execute the following method: Perform a time series change analysis on the process data collection results, perform a time series prediction of the production status based on the time series change analysis results, and establish a time series prediction result; perform data segmentation based on the process data collection results and the time series prediction results, and generate a backtracking compensation weight and a predictive compensation weight for the target corresponding to the static weight group based on the data segmentation result; perform corresponding static weight group compensation based on the backtracking compensation weight and the predictive compensation weight, and construct a real-time dynamic weight with the compensation result.

[0049] Furthermore, the updating module 14 is used to execute the following method: Establish a basic objective function, which includes a production efficiency function, a mold life function, a material utilization function, and a product quality function; weighted fit the basic objective function through the real-time dynamic weight to establish a multi-objective evaluation function; perform a weight change rate analysis on the real-time dynamic weight, and establish an optimization priority based on the change value of the weight change rate.

[0050] Furthermore, the optimization module 15 is used to execute the following method: An initial solution space is established, wherein the initial solution space contains initial solutions of a first proportion and initial solutions of a second proportion, wherein the initial solutions of the first proportion are solution sets constructed by historical parameters, and the initial solutions of the second proportion are random solution sets, and the first proportion is greater than the second proportion, and the sum of the first proportion and the second proportion is 1, and each solution in the initial solution space represents a parameter combination of multiple objectives; the fitness of the solutions in the initial solution space is calculated using the multi-objective evaluation function, and a leader solution set is established according to the fitness calculation result, wherein the leader solution set includes at least one historical parameter solution and one random solution; the remaining solutions in the initial solution space are searched and updated based on the leader solution and the optimization priority, and the leader solution set is updated, and the initial solution space is reconstructed with the updated leader solution set and the search update result; a search iteration is performed, and the adjustment optimization is completed according to the search iteration result.

[0051] Furthermore, the optimization module 15 is used to execute the following method: An iterative observation layer is established, and the iterative observation layer is used to observe and evaluate the results of each round of search iterations; when the observation and evaluation results meet a preset trigger threshold, a leader solution set constraint instruction is generated, and the preset trigger threshold is a threshold value where the number of iterations is within a preset number range and the search variability is lower than a preset variability value; the leader solution set generation constraint for the next round of iteration is performed according to the leader solution set constraint instruction.

[0052] Furthermore, the optimization module 15 is used to execute the following method: Perform production monitoring and establish a real-time response database; perform response evaluation on the real-time response database and establish a reinforcement learning controller; and use the reinforcement learning controller to update and iterate a multi-objective evaluation function.

[0053] Furthermore, the optimization module 15 is used to execute the following method: Establish a multi-level warning mapping set; use the real-time response database to perform warning trigger analysis on the multi-level warning mapping set; and issue a warning alert based on the mapping warning level corresponding to the warning trigger analysis result.

[0054] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0055] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0056] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. Mold parameter adjustment method based on multi-objective optimization algorithm, It is characterized in that The method comprises: Access the management equipment of the factory, call the production database, and divide the production status based on the production database to establish the production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target; Utilize the sensor network to perform process data collection, and synchronize the process data collection results to the data processing center through the MQTT protocol, and the data processing center stores the production status division results; The data processing center matches the process data collection results with the production status division results, and constructs real-time dynamic weights with the static weight group corresponding to the matching results; Use real-time dynamic weights to update multi-objective evaluation functions and establish configuration optimization priorities; The mold parameters are adjusted and optimized according to the multi-objective evaluation function and the optimization priority, and the mold parameters are optimized based on the adjustment and optimization results.

2. The mold parameter adjustment method based on the multi-objective optimization algorithm according to claim 1, It is characterized in that The method of constructing the real-time dynamic weights with the static weights corresponding to the matching results also includes: Performing a time series change analysis on the process data collection results, performing a production status time series prediction based on the time series change analysis results, and establishing a time series prediction result; Perform data segmentation according to the process data collection results and the time series prediction results, and generate the backtracking compensation weight and the prediction compensation weight of the static weight group corresponding to the target according to the data segmentation results; The corresponding static weight group compensation is carried out according to the retrospective compensation weight and the predicted compensation weight, and the real-time dynamic weight is constructed based on the compensation results.

3. The mold parameter adjustment method based on the multi-objective optimization algorithm according to claim 1, It is characterized in that The method of updating the multi-objective evaluation function using real-time dynamic weights and establishing configuration optimization priorities also includes: Establishing basic objective functions, wherein the basic objective functions include production efficiency function, mold life function, material utilization function, and product quality function; The basic objective function is weightedly fitted by the real-time dynamic weight to establish a multi-objective evaluation function; The weight change rate of the real-time dynamic weight is analyzed, and the optimization priority is established according to the change value of the weight change rate.

4. The mold parameter adjustment method based on the multi-objective optimization algorithm as claimed in claim 3, It is characterized in that The step of adjusting and optimizing the mold parameters according to the multi-objective evaluation function and the optimization priority also includes: Establishing an initial solution space, wherein the initial solution space includes an initial solution of a first proportion and an initial solution of a second proportion, wherein the initial solution of the first proportion is a solution set constructed by historical parameters, and the initial solution of the second proportion is a random solution set, and the first proportion is greater than the second proportion, and the sum of the first proportion and the second proportion is 1, and each solution in the initial solution space represents a parameter combination of multiple objectives; Using the multi-objective evaluation function to calculate the fitness of the solutions in the initial solution space, and establishing a leader solution set according to the fitness calculation results, wherein the leader solution set includes at least one historical parameter solution and one random solution; The remaining solutions in the initial solution space are searched and updated based on the leader solution and the optimization priority, and the leader solution set is updated, and the initial solution space is reconstructed with the updated leader solution set and the search update results; Execute search iterations and complete optimization adjustments based on the search iteration results.

5. The mold parameter adjustment method based on the multi-objective optimization algorithm according to claim 4, It is characterized in that The performing of the search iteration further comprises: Establishing an iterative observation layer, wherein the iterative observation layer is used to observe and evaluate the results of each round of search iteration; When the observation evaluation result meets the preset trigger threshold, a leader solution set constraint instruction is generated, wherein the preset trigger threshold is a threshold value where the number of iterations is within a preset number range and the search anomaly is lower than a preset anomaly value; The leader solution set generation constraints for the next round of iteration are performed according to the leader solution set constraint instructions.

6. The mold parameter adjustment method based on the multi-objective optimization algorithm according to claim 1, It is characterized in that The method further comprises: Perform production monitoring and establish a real-time response database; Performing response evaluation on the real-time response database and establishing a reinforcement learning controller; The reinforcement learning controller is used to update and iterate the multi-objective evaluation function.

7. The mold parameter adjustment method based on the multi-objective optimization algorithm according to claim 6, It is characterized in that The method further comprises: Establish a multi-level warning mapping set; Use the real-time response database to perform warning trigger analysis on a multi-level warning mapping set; The early warning is issued according to the mapped early warning level corresponding to the early warning trigger analysis result.

8. Mould parameter adjustment system based on multi-objective optimization algorithm, It is characterized in that A system for implementing the mold parameter adjustment method based on a multi-objective optimization algorithm according to any one of claims 1 to 7, the system comprising: A production status division module, which is used to access the management equipment of the factory, call the production database, and perform production status division based on the production database to establish a production status division result, wherein each production status division result corresponds to a static weight group, and the static weight group is a set of weight proportions reflecting the production target; A data acquisition module, wherein the data acquisition module is used to perform process data acquisition using a sensor network and synchronize the process data acquisition results to a data processing center through the MQTT protocol, wherein the data processing center stores the production status division results; A matching module, the matching module is used to match the process data collection results with the production status division results by the data processing center, and construct a real-time dynamic weight with a static weight group corresponding to the matching result; An update module, the update module is used to update the multi-objective evaluation function using real-time dynamic weights and establish a configuration optimization priority; The optimization module is used to adjust and optimize the mold parameters according to the multi-objective evaluation function and the optimization priority, and optimize the mold parameters based on the adjustment optimization result.

Images