Intelligent production management system based on interface agent performance test data

Abstract

The invention relates to the technical field of chemical production management, in particular to an intelligent production management system based on interface agent performance test data, and aims to solve the problems that key performance data of an interface agent cannot be comprehensively acquired, dynamic monitoring and quantitative analysis of the overall performance of the interface agent cannot be realized, the consistency among batches cannot be effectively evaluated, and the production efficiency cannot be improved in the prior art. And the stable product quality cannot be ensured. According to the method, key performance data of the interface agent are comprehensively collected, the change rate of the indexes is calculated, dynamic monitoring and quantitative analysis of the overall performance of the interface agent are achieved, the overall performance evaluation coefficient ZXP is calculated through a weighted comprehensive evaluation method, the product performance is rapidly judged in combination with the preset threshold value, and the product quality is improved. And meanwhile, the average value, the standard deviation and the variable coefficient of the key performance indexes of each batch are calculated, so that the consistency between the batches is effectively evaluated, and the stable product quality is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of chemical production management, and more specifically, to an intelligent production management system based on interface agent performance test data. Background Art

[0002] As an important type of chemical material, the performance of the interface agent directly affects the use effect and service life of related products. Traditional interface agent production management mainly relies on manual experience and regular spot checks. This method has lag, making it difficult to detect problems in the production process in a timely manner and unable to adjust production parameters in real time.

[0003] The patent application with the publication number CN118569616A discloses an intelligent production management method and system based on interface agent performance test data. The present invention first obtains production data, test performance data, and test environment data; further reduces the dimension of the production data and obtains the importance degree parameters of each main dimension; further obtains the first outlier degree of each interface agent sample; further obtains the second outlier degree of each interface agent sample; further obtains the third outlier degree of each interface agent sample; further obtains an optimized multiple linear regression model; and finally conducts intelligent production management on the interface agent. By analyzing the outlier similarity between production data and test performance data and combining the outlier abnormal characteristics of test environment data, the present invention enables the regression model to avoid the problem of sample deviation, improves the fitting accuracy of the regression model, thereby conducts intelligent production management on the interface agent, improves production efficiency, and reduces production costs. However, the above reference patent analyzes the outlier similarity between production and test data and the abnormal characteristics of the test environment, adopts non-linear dimensionality reduction and optimizes the regression model to achieve intelligent management and efficiency improvement of interface agent production. However, it cannot comprehensively collect the key performance data of the interface agent, cannot achieve dynamic monitoring and quantitative analysis of its overall performance, cannot effectively evaluate the consistency between batches, cannot ensure stable product quality, cannot conduct dynamic quality monitoring on the interface agent production process, cannot automatically identify the production quality status, and at the same time cannot predict the key performance changes of the interface agent based on production parameters, cannot predict product performance in advance, and cannot quickly optimize parameters, reducing the controllability of production and product quality.

[0004] Therefore, we propose an intelligent production management system based on interface agent performance test data for the above problems. Summary of the Invention

[0005] The object of the present invention is to provide an intelligent production management system based on the performance test data of interface agents, which solves the problems in the prior art that the key performance data of interface agents cannot be comprehensively collected, the dynamic monitoring and quantitative analysis of their overall performance cannot be achieved, the consistency between batches cannot be effectively evaluated, the product quality cannot be guaranteed to be stable, the dynamic quality monitoring of the production process of interface agents cannot be carried out, the production quality status cannot be automatically identified, and at the same time, the key performance changes of interface agents cannot be predicted based on production parameters, the product performance cannot be predicted in advance, and the parameters cannot be quickly optimized, reducing the controllability of production and the product quality.

[0006] The object of the present invention is achieved by the following technical solutions: An intelligent production management system based on the performance test data of interface agents, which is applied to a production management platform, includes: A data acquisition module, which is used to acquire the performance test data of interface agents and perform preprocessing operations on the acquired performance test data of interface agents; A performance index evaluation module, which is used to calculate the key performance indexes of interface agents according to the acquired performance test data of interface agents and evaluate the overall performance of interface agents and the consistency between batches; A quality warning module, which is used to monitor the quality of the production process of interface agents and timely warn of potential quality problems; A production parameter optimization module, which based on the historical performance test data of interface agents and historical production parameters, establishes a prediction model between the key performance indexes of interface agents and production parameters, and uses the established prediction model to optimize production parameters.

[0007] As a preferred embodiment of the present invention, the specific process of the performance index evaluation module evaluating the overall performance of interface agents and the consistency between batches is as follows: Obtain the performance test data of interface agents. The performance test data of interface agents includes surface tension, contact angle, viscosity, pH value, drying time, conductivity, adhesion, and solubility. Generate an acquisition period, and the acquisition period duration is set to one year. Calculate the key performance indexes of interface agents according to the acquired performance test data of interface agents. The key performance indexes of interface agents include surface tension change rate, viscosity change rate, pH value change rate, drying time change rate, and adhesion change rate; The specific steps for calculating the surface tension change rate are as follows: Obtain the surface tension value γ(t) of the interface agent at each time point t, and use the difference formula to calculate the surface tension change amount between adjacent time points: ∆γ(t)=γ(t)-γ(t - 1), where γ(t) represents the surface tension at time point t, and γ(t - 1) represents the surface tension at time point t - 1; Use the following formula to calculate the change rate BZ(t) of the surface tension at time point t: 。

[0008] As a preferred embodiment of the present invention, the method of calculating the surface tension change rate BZ(t) can be similarly used to obtain the viscosity change rate ND(t), the pH value change rate PH(t), the drying time change rate GS(t), and the adhesion change rate FL(t). The overall performance evaluation coefficient ZXP is calculated by the following formula: ZXP = c1*BZ(t) + c2*ND(t) + c3*PH(t) + c4*GS(t) + c5*FL(t), where c1, c2, c3, c4, and c5 are all weighting coefficients. The overall performance evaluation coefficient ZXP is compared with the preset overall performance evaluation coefficient threshold: If the overall performance evaluation coefficient ZXP is less than the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is good; If the overall performance evaluation coefficient ZXP is greater than or equal to the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is poor.

[0009] As a preferred embodiment of the present invention, the performance test data of the interface agent for each batch is obtained. Assuming there are n batches, the average surface tension of the interface agent is calculated by the following formula : , where γ i is the surface tension value of the interface agent in the i-th batch; The standard deviation s of the surface tension of the interface agent is calculated using the following formula γ : ; The coefficient of variation CV of the surface tension of the interface agent is calculated by the following formula γ : ; The coefficient of variation CV of the surface tension γ is compared with the preset coefficient of variation threshold of the surface tension; If the coefficient of variation CV of the surface tension γ is less than the preset coefficient of variation threshold of the surface tension, it indicates that the consistency of the surface tension between batches is qualified; If the coefficient of variation CV of the surface tension γ is greater than or equal to the preset coefficient of variation threshold of the surface tension, it indicates that the consistency of the surface tension between batches is unqualified; The method of using the batch - to - batch consistency of the determined surface tension to judge whether it is qualified can similarly obtain whether the batch consistency of viscosity, pH value, drying time, and adhesion is qualified. If the batch consistency of surface tension, viscosity, pH value, drying time, and adhesion are all qualified, it indicates that the consistency among n batches of the interface agent is qualified; otherwise, it indicates that the consistency among n batches of the interface agent is unqualified.

[0010] As a preferred embodiment of the present invention, the specific process of the quality warning module for quality monitoring during the production process of the interface agent is as follows: Obtain the historical key process parameters during the production process of the interface agent. The key process parameters include reaction temperature, reaction pressure, stirring speed, and reaction time, and the production monitoring period. Divide the monitoring period into multiple monitoring time periods; Obtain the reaction temperature of the interface agent during multiple monitoring time periods, and calculate the arithmetic mean of the obtained multiple reaction temperatures. Denote the arithmetic mean of the multiple reaction temperatures as the average reaction temperature PFW; Using the method of calculating the average reaction temperature PFW of the interface agent during the production process, similarly, the average reaction pressure PFY, average stirring speed PJS, and average reaction time PFS can be obtained.

[0011] As a preferred embodiment of the present invention, obtain the key performance indicators of the interface agent calculated based on historical interface agent performance test data, and combine the average reaction temperature PFW, average reaction pressure PFY, average stirring speed PJS, average reaction time PFS, viscosity change rate ND(t), pH value change rate PH(t), drying time change rate GS(t), and adhesion change rate FL(t) to construct a quality monitoring feature matrix ZJT; Use the constructed feature matrix ZJT as the input of the machine learning model, and use the label vector w as the output of the machine learning model. The label vector w represents the production quality status of the interface agent. The output of the label vector w is 0, 1, or 2. 0 indicates that the production quality of the interface agent is excellent, 1 indicates that the production quality of the interface agent is normal, and 2 indicates that the production quality of the interface agent is abnormal. Using the label vector w as the prediction target and minimizing the sum of the prediction errors of all training data as the training target, train the machine learning model until the sum of the prediction errors reaches convergence and then stop training to obtain the machine learning model that predicts the label vector w.

[0012] As a preferred embodiment of the present invention, obtain the real - time key process parameters and real - time interface agent performance test data during the production process of the interface agent, process them to construct a real - time quality monitoring feature matrix ZJT, and perform quality monitoring on the production process of the interface agent through the trained machine learning model; If the label vector w output by the machine learning model is 0, it indicates that the production quality of the interface agent is excellent; If the label vector w output by the machine learning model is 1, it indicates that the production quality of the interface agent is normal; If the label vector w output by the machine learning model is 2, it indicates that the production quality of the interface agent is abnormal, generates an abnormal warning signal and sends it to the production management platform; After receiving the abnormal warning signal, the production management platform immediately sends a warning notice to the management personnel and immediately takes corresponding measures to regulate the production process.

[0013] As a preferred embodiment of the present invention, the specific process of the production parameter optimization module establishing a prediction model between the key performance indicators of the interface agent and the production parameters is as follows: Collect historical interface agent performance test data and historical production parameters. The historical production parameters include reaction temperature, stirring speed, reaction time, and the proportion of key additives, generate a collection period, and the collection period duration is set to one month; Using the method of calculating the surface tension change rate BZ(t), the reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB, and key additive proportion change rate TBB can be obtained in the same way. Combine the reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB, key additive proportion change rate TBB, and historical interface agent performance test data to construct a prediction matrix GXJ of key performance indicator changes.

[0014] As a preferred embodiment of the present invention, take the prediction matrix GXJ as the input of the machine learning model, and take the key performance indicator change matrix of the interface agent produced in a future period corresponding to each prediction matrix GXJ as the output of the machine learning model. Take the key performance indicator change matrix of the interface agent produced in a future period as the prediction target, and take minimizing the sum of prediction errors of all training data as the training target. Train the machine learning model until the sum of prediction errors reaches convergence and then stop training to obtain a prediction model for the change of key performance indicators of the interface agent. The expression formula of the prediction model for the change of key performance indicators of the interface agent is as follows: ; Where GXB represents the key performance indicator change matrix of the interface agent produced in a future period, β1, β2, β3, and β4 are all regression coefficients, λ is a random error term, ∆γ represents the change value of surface tension, ∆η represents the change value of viscosity, ∆p represents the change value of PH value, ∆g represents the change value of drying time, and ∆f represents the change value of adhesion.

[0015] As a preferred embodiment of the present invention, real-time interface agent performance test data and real-time production parameters are obtained, converted into a corresponding prediction matrix GXJ, and input into the prediction model for changes in the key performance indicators of the interface agent. Through the model, a real-time key performance indicator change matrix GXB of the interface agent produced in the next period of time is obtained, and the obtained real-time key performance indicator change matrix GXB of the interface agent is sent to the production management platform; After receiving the real-time key performance indicator change matrix GXB of the interface agent, the production management platform immediately takes adjustment measures to optimize the production parameters in the production process of the interface agent.

[0016] Compared with the prior art, the advantages of the present invention are as follows: (1) In the present invention, the performance index evaluation module realizes the dynamic monitoring and quantitative analysis of the overall performance by comprehensively collecting the key performance data of the interface agent and calculating the change rates of these indicators. The overall performance evaluation coefficient ZXP is calculated by using the weighted comprehensive evaluation method, and the quality of the product is quickly judged in combination with the preset threshold. At the same time, by calculating the average value, standard deviation and coefficient of variation of the key performance indicators of each batch, the consistency between batches is effectively evaluated, ensuring the stable product quality; (2) In the present invention, the quality early warning module uses a machine learning model to dynamically monitor the quality of the interface agent production process, integrates key parameters such as reaction temperature, pressure, stirring speed, and performance indicators such as viscosity change rate and PH value change rate. It can construct a feature matrix in real time, automatically identify the production quality status, and quickly issue an early warning when an abnormality is detected. Through continuous data analysis, the production process is optimized, and the product consistency and quality stability are improved; (3) In the present invention, the production parameter optimization module uses historical data and a machine learning model to predict the key performance changes of the interface agent based on production parameters such as reaction temperature and stirring speed. By inputting real-time data, the product performance can be predicted in advance, enabling the production management platform to quickly optimize the parameters, improve the product quality and consistency. This module is convenient for adjustment according to the actual production situation, enhances the efficiency of process optimization, improves the controllability of production and product quality, and reduces resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is the system block diagram of Embodiment 1 in the present invention; Figure 2 It is the system block diagram of Embodiment 2 in the present invention; Figure 3 It is the logic flow schematic diagram in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will combine the accompanying drawings in the embodiments of the present invention; and clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Embodiment 1: As Figure 1 and Figure 3 shown, the intelligent production management system based on the interface agent performance test data proposed by the present invention is applied to the production management platform and includes: A data acquisition module, which is used to collect the interface agent performance test data and perform preprocessing operations on the collected interface agent performance test data. The preprocessing operations include but are not limited to data cleaning, data conversion, and data integration; The data acquisition module improves data quality by efficiently collecting and preprocessing the interface agent performance data, supports real-time monitoring of key indicators, quickly identifies anomalies and optimizes processes. It integrates multi-source data, provides visual reports and automated warnings, ensures timely decision-making and problem prevention. Long-term data accumulation promotes process optimization, improves efficiency, reduces costs, and realizes intelligent production management.

[0020] A performance index evaluation module, which is used to calculate the key performance indicators of the interface agent according to the collected interface agent performance test data and evaluate the overall performance of the interface agent and the consistency between batches; The specific process of the performance index evaluation module evaluating the overall performance of the interface agent and the consistency between batches is as follows: Obtain the interface agent performance test data. The interface agent performance test data includes surface tension, contact angle, viscosity, pH value, drying time, conductivity, adhesion, and solubility. Generate a collection period, and the collection period duration is set to one year. Calculate the key performance indicators of the interface agent according to the collected interface agent performance test data. The key performance indicators of the interface agent include surface tension change rate, viscosity change rate, pH value change rate, drying time change rate, and adhesion change rate; The specific steps for calculating the surface tension change rate are as follows: Obtain the surface tension value γ(t) of the interface agent at each time point t, and use the difference formula to calculate the surface tension change amount between adjacent time points: ∆γ(t)=γ(t)-γ(t - 1), where γ(t) represents the surface tension at time point t, and γ(t - 1) represents the surface tension at time point t - 1; Use the following formula to calculate the change rate BZ(t) of the surface tension at time point t: ; By using the method of calculating the surface tension change rate BZ(t), the viscosity change rate ND(t), the pH value change rate PH(t), the drying time change rate GS(t), and the adhesion change rate FL(t) can be obtained in the same way. The overall performance evaluation coefficient ZXP is calculated by the following formula: ZXP = c1 * BZ(t) + c2 * ND(t) + c3 * PH(t) + c4 * GS(t) + c5 * FL(t), where c1, c2, c3, c4, and c5 are all weight coefficients. The overall performance evaluation coefficient ZXP is compared with the preset overall performance evaluation coefficient threshold: If the overall performance evaluation coefficient ZXP is less than the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is good; If the overall performance evaluation coefficient ZXP is greater than or equal to the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is poor; Obtain the performance test data of the interface agent for each batch. Suppose there are n batches. The average surface tension of the interface agent is calculated by the following formula : , where γ i is the surface tension value of the interface agent in the i-th batch; The standard deviation s of the surface tension of the interface agent is calculated by the following formula γ : ; The coefficient of variation CV of the surface tension of the interface agent is calculated by the following formula γ : ; The coefficient of variation CV of the surface tension γ is compared with the preset coefficient of variation threshold of the surface tension; If the coefficient of variation CV of the surface tension γ is less than the preset coefficient of variation threshold of the surface tension, it indicates that the consistency of the surface tension between batches is qualified; If the coefficient of variation CV of the surface tension γ is greater than or equal to the preset coefficient of variation threshold of the surface tension, it indicates that the consistency of the surface tension between batches is unqualified; By using the method of determining whether the consistency of the surface tension between batches is qualified, it can be obtained in the same way whether the batch consistency of viscosity, pH value, drying time, and adhesion is qualified. If the batch consistency of surface tension, viscosity, pH value, drying time, and adhesion is all qualified, it indicates that the consistency of the n batches of the interface agent is qualified. Otherwise, it indicates that the consistency of the n batches of the interface agent is unqualified; ​​​​​The performance index evaluation module realizes the dynamic monitoring and quantitative analysis of the overall performance of the interface agent by comprehensively collecting the key performance data of the interface agent (such as surface tension, viscosity, pH value, etc.) and calculating the change rates of these indexes. The overall performance evaluation coefficient ZXP is calculated by using the weighted comprehensive evaluation method, and the quality of the product is quickly judged by combining the preset threshold. At the same time, the consistency between batches is effectively evaluated by calculating the average value, standard deviation and coefficient of variation of the key performance indexes of each batch, ensuring the stable product quality.

[0021] The quality early warning module is used to monitor the quality of the interface agent production process and timely warn of potential quality problems; The specific process of the quality early warning module monitoring the quality of the interface agent production process is as follows: Obtain the historical key process parameters in the interface agent production process. The key process parameters include reaction temperature, reaction pressure, stirring speed and reaction time, and the production monitoring period. Divide the monitoring period into multiple monitoring time periods; Obtain the reaction temperature of the interface agent in multiple monitoring time periods, and calculate the arithmetic mean of the obtained multiple reaction temperatures. Denote the arithmetic mean of the multiple reaction temperatures as the average reaction temperature PFW; Using the method of calculating the average reaction temperature PFW of the interface agent in the production process, the average reaction pressure PFY, the average stirring speed PJS and the average reaction time PFS can be obtained in the same way; Obtain the key performance indexes of the interface agent calculated according to the historical interface agent performance test data. Combine the average reaction temperature PFW, the average reaction pressure PFY, the average stirring speed PJS, the average reaction time PFS, the viscosity change rate ND(t), the pH value change rate PH(t), the drying time change rate GS(t) and the adhesion change rate FL(t) to construct the quality monitoring feature matrix ZJT; Take the constructed feature matrix ZJT as the input of the machine learning model, and take the label vector w as the output of the machine learning model. The label vector w represents the production quality status of the interface agent. The output of the label vector w is 0, 1 or 2. 0 indicates that the production quality of the interface agent is excellent, 1 indicates that the production quality of the interface agent is normal, and 2 indicates that the production quality of the interface agent is abnormal. Take the label vector w as the prediction target, and take minimizing the sum of the prediction errors of all training data as the training target. Train the machine learning model until the sum of the prediction errors reaches convergence and then stop training to obtain the machine learning model that predicts the label vector w; Obtain the real-time key process parameters and real-time interface agent performance test data in the interface agent production process, process them and construct the real-time quality monitoring feature matrix ZJT, and monitor the quality of the interface agent production process through the trained machine learning model; If the label vector w output by the machine learning model is 0, it indicates that the production quality of the interface agent is excellent; If the label vector w output by the machine learning model is 1, it indicates that the production quality of the interface agent is normal; If the label vector w output by the machine learning model is 2, it indicates that the production quality of the interface agent is abnormal, and an abnormal warning signal is generated and sent to the production management platform; After receiving the abnormal warning signal, the production management platform immediately sends a warning notification to the management personnel and immediately takes corresponding measures to regulate the production process; The quality warning module uses a machine learning model to dynamically monitor the quality of the interface agent production process, integrating key parameters such as reaction temperature, pressure, stirring speed, and performance indicators such as viscosity change rate and pH value change rate. It can construct a feature matrix in real time, automatically identify the production quality status (excellent, normal or abnormal), and quickly issue a warning when an abnormality is detected. The module optimizes the production process through continuous data analysis and improves product consistency and quality stability.

[0022] Embodiment 2: The technical solution of the embodiment of the present invention is different from that of embodiment 1 in that: like Figure 2 As shown, the production parameter optimization module establishes a prediction model between the key performance indicators of the interface agent and the production parameters based on the historical interface agent performance test data and historical production parameters, and optimizes the production parameters using the established prediction model; The specific process of the production parameter optimization module to establish a prediction model between the key performance indicators of the interface agent and the production parameters is as follows: Collect historical interface agent performance test data and historical production parameters, including reaction temperature, stirring speed, reaction time and key additive ratio, and generate a collection cycle. The collection cycle duration is set to one month. By using the method of calculating the surface tension change rate BZ(t), the reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB and key additive ratio change rate TBB can be obtained. The reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB, key additive ratio change rate TBB and historical interface agent performance test data are combined to construct a prediction matrix GXJ for key performance indicator changes. Use the prediction matrix GXJ as the input of the machine learning model, and use the change matrix of the key performance indicators of the interface agent produced in a future period corresponding to each group of prediction matrices GXJ as the output of the machine learning model. Take the change matrix of the key performance indicators of the interface agent produced in a future period as the prediction target, and use minimizing the sum of the prediction errors of all training data as the training target to train the machine learning model until the sum of the prediction errors converges and then stop training to obtain the prediction model for the change of the key performance indicators of the interface agent. The expression formula of the prediction model for the change of the key performance indicators of the interface agent is as follows: ; Where GXB represents the change matrix of the key performance indicators of the interface agent produced in a future period, β1, β2, β3, and β4 are all regression coefficients, λ is a random error term, ∆γ represents the change value of the surface tension, ∆η represents the change value of the viscosity, ∆p represents the change value of the PH value, ∆g represents the change value of the drying time, and ∆f represents the change value of the adhesion; Obtain the real-time interface agent performance test data and real-time production parameters, convert them into the corresponding prediction matrix GXJ and input it into the prediction model for the change of the key performance indicators of the interface agent. Through the model, obtain the real-time change matrix GXB of the key performance indicators of the interface agent produced in a future period, and send the obtained real-time change matrix GXB of the key performance indicators of the interface agent to the production management platform; After receiving the real-time change matrix GXB of the key performance indicators of the interface agent, the production management platform immediately takes corresponding measures to optimize the production parameters in the production process of the interface agent. The specific content of the adjustment measures is as follows: If ∆γ > 0, then reduce the reaction temperature, reduce the proportion of key additives or extend the reaction time. Otherwise, increase the reaction temperature, increase the proportion of key additives or shorten the reaction time; If ∆η > 0, then reduce the reaction temperature, reduce the stirring speed or adjust the additive proportion. Otherwise, increase the reaction temperature, increase the stirring speed or adjust the additive proportion; If ∆p > 0, then add acidic substances. Otherwise, add alkaline substances; If ∆g > 0, then increase the reaction temperature, increase the stirring speed or adjust the additive proportion. Otherwise, reduce the reaction temperature, reduce the stirring speed or adjust the additive proportion; If ∆f > 0, then finely adjust the temperature, stirring speed or additive proportion to reduce the adhesion. Otherwise, finely adjust the temperature, stirring speed or additive proportion to increase the adhesion; The production parameter optimization module uses historical data to build a prediction model, accurately predicts the performance of the interface agent, and adjusts production parameters according to real-time data to ensure stable product performance. It comprehensively considers factors such as reaction temperature, stirring speed, and additive ratio, provides specific adjustment strategies (such as for changes in surface tension, viscosity, pH value, etc.), improves product quality, reduces experimental errors, improves efficiency and resource utilization through precise control and timely adjustment. The adjustment strategy based on the prediction results provides strong support for management decisions, enhances product competitiveness and production controllability, and significantly improves production efficiency and quality.

[0023] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and its improved concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent production management system based on interface agent performance test data is applied to the production management platform, which is characterized by: include: The data acquisition module is used to collect interface agent performance test data and perform preprocessing operations on the collected interface agent performance test data; The performance index evaluation module is used to calculate the key performance indicators of the interface agent based on the collected interface agent performance test data, and evaluate the overall performance of the interface agent and the consistency between batches; The quality early warning module is used to monitor the quality of the interface agent production process and promptly warn of potential quality problems; The production parameter optimization module establishes a prediction model between the key performance indicators of the interface agent and the production parameters based on the historical interface agent performance test data and historical production parameters, and optimizes the production parameters using the established prediction model.

2. The intelligent production management system based on interface agent performance test data according to claim 1 is characterized in that: The specific process of calculating the key performance indicators of the interface agent by the performance indicator evaluation module is as follows: Acquire the performance test data of the interface agent, including surface tension, contact angle, viscosity, pH value, drying time, conductivity, adhesion and solubility, generate a collection cycle, and set the collection cycle duration to one year. Calculate the key performance indicators of the interface agent based on the collected performance test data of the interface agent, including the surface tension change rate, viscosity change rate, pH value change rate, drying time change rate and adhesion change rate; The specific steps for calculating the rate of change of surface tension are as follows: Obtain the surface tension value γ(t) of the interfacial agent at each time point t, and use the difference formula to calculate the change in surface tension between adjacent time points: ∆γ(t)=γ(t)-γ(t-1), where γ(t) represents the surface tension at time point t, and γ(t-1) represents the surface tension at time point t-1; The rate of change of surface tension at time point t, BZ(t), is calculated using the following formula: 。 3. The intelligent production management system based on interface agent performance test data according to claim 2 is characterized in that: The specific process of the performance index evaluation module evaluating the overall performance of the interface agent is as follows: Using the method of calculating the surface tension change rate BZ(t), the viscosity change rate ND(t), pH value change rate PH(t), drying time change rate GS(t) and adhesion change rate FL(t) can be obtained in the same way, and the overall performance evaluation coefficient ZXP can be calculated by the following formula: ZXP=c1*BZ(t)+c2*ND(t)+c3*PH(t)+c4*GS(t)+c5*FL(t), where c1, c2, c3, c4 and c5 are weight coefficients. The overall performance evaluation coefficient ZXP is compared with the preset overall performance evaluation coefficient threshold: If the overall performance evaluation coefficient ZXP is less than the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is good; If the overall performance evaluation coefficient ZXP is greater than or equal to the preset overall performance evaluation coefficient threshold, it indicates that the overall performance of the interface agent is poor.

4. The intelligent production management system based on interface agent performance test data according to claim 3 is characterized in that: The specific process of the performance index evaluation module evaluating the consistency between batches of the interface agent is as follows: Obtain the performance test data of each batch of the interface agent. Assuming there are n batches, calculate the average surface tension of the interface agent using the following formula: : , where γ i is the surface tension value of the interfacial agent in the i-th batch; The standard deviation of the surface tension of the interfacial agent is calculated using the following formula: γ : ； The coefficient of variation of the surface tension of the interfacial agent CV is calculated by the following formula γ : ； The coefficient of variation of surface tension CV γ Compare with the preset surface tension coefficient of variation threshold; If the coefficient of variation of surface tension CV γ If the value is less than the preset surface tension coefficient of variation threshold, it indicates that the batch-to-batch consistency of the surface tension is qualified; If the coefficient of variation of surface tension CV γ If it is greater than or equal to the preset surface tension coefficient of variation threshold, it indicates that the batch-to-batch consistency of surface tension is unqualified; The method of determining whether the batch consistency of surface tension is qualified can be used to determine whether the batch consistency of viscosity, pH value, drying time and adhesion is qualified. If the batch consistency of surface tension, viscosity, pH value, drying time and adhesion are all qualified, it means that the consistency of n batches of the interface agent is qualified. Otherwise, it means that the consistency of n batches of the interface agent is unqualified.

5. The intelligent production management system based on interface agent performance test data according to claim 1 is characterized in that: The specific process of the quality early warning module processing the key process parameters in the production process of the interface agent is as follows: Obtain the historical key process parameters in the production process of the interface agent, including reaction temperature, reaction pressure, stirring speed and reaction time, production monitoring cycle, and divide the monitoring cycle into multiple monitoring periods; Obtaining the reaction temperature of the interfacial agent in multiple monitoring periods, calculating the arithmetic average of the obtained multiple reaction temperatures, and recording the arithmetic average of the multiple reaction temperatures as the average reaction temperature PFW; The average reaction pressure PFY, the average stirring speed PJS and the average reaction time PFS can be obtained by calculating the average reaction temperature PFW of the interface agent during the production process.

6. The intelligent production management system based on interface agent performance test data according to claim 5 is characterized in that: The specific process of the quality early warning module to monitor the quality of the interface agent production process is as follows: The key performance indicators of the interface agent calculated based on the historical interface agent performance test data are obtained, and the average reaction temperature PFW, the average reaction pressure PFY, the average stirring speed PJS, the average reaction time PFS, the viscosity change rate ND(t), the pH value change rate PH(t), the drying time change rate GS(t) and the adhesion change rate FL(t) are combined to construct the quality monitoring feature matrix ZJT; The constructed feature matrix ZJT is used as the input of the machine learning model, and the label vector w is used as the output of the machine learning model. The label vector w represents the production quality status of the interface agent. The output of the label vector w is 0, 1 or 2. 0 indicates that the production quality of the interface agent is excellent, 1 indicates that the production quality of the interface agent is normal, and 2 indicates that the production quality of the interface agent is abnormal. The label vector w is used as the prediction target, and minimizing the sum of prediction errors of all training data is used as the training target. The machine learning model is trained until the sum of prediction errors converges and the training is stopped to obtain a machine learning model that predicts the label vector w.

7. The intelligent production management system based on interface agent performance test data according to claim 6 is characterized in that: The specific process of the quality warning module to timely warn of potential quality problems is as follows: Obtain the real-time key process parameters and real-time interface agent performance test data in the interface agent production process, construct the real-time quality monitoring feature matrix ZJT after processing, and monitor the quality of the interface agent production process through the trained machine learning model; If the label vector w output by the machine learning model is 0, it indicates that the production quality of the interface agent is excellent; If the label vector w output by the machine learning model is 1, it indicates that the production quality of the interface agent is normal; If the label vector w output by the machine learning model is 2, it indicates that the production quality of the interface agent is abnormal, and an abnormal warning signal is generated and sent to the production management platform; After receiving the abnormal warning signal, the production management platform immediately sends a warning notification to the management personnel and immediately takes corresponding measures to regulate the production process.

8. The intelligent production management system based on interface agent performance test data according to claim 1 is characterized in that: The specific process of the production parameter optimization module processing the historical interface agent performance test data and historical production parameters is as follows: Collect historical interface agent performance test data and historical production parameters, including reaction temperature, stirring speed, reaction time and key additive ratio, and generate a collection cycle. The collection cycle duration is set to one month. By using the method of calculating the surface tension change rate BZ(t), we can also obtain the reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB and key additive ratio change rate TBB. The reaction temperature change rate FWB, stirring speed change rate JSB, reaction time change rate FSB, key additive ratio change rate TBB and historical interface agent performance test data are combined to construct a prediction matrix GXJ for changes in key performance indicators.

9. The intelligent production management system based on interface agent performance test data according to claim 8, characterized in that: The specific process of the production parameter optimization module to establish a prediction model between the key performance indicators of the interface agent and the production parameters is as follows: The prediction matrix GXJ is used as the input of the machine learning model, and the key performance indicator change matrix of the interface agent produced in the future corresponding to each group of prediction matrices GXJ is used as the output of the machine learning model. The key performance indicator change matrix of the interface agent produced in the future is used as the prediction target, and the sum of the prediction errors of all training data is minimized as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence and the training is stopped. The prediction model of the key performance indicator change of the interface agent is obtained. The key performance indicator change prediction model of the interface agent is expressed as follows: ； Among them, GXB represents the change matrix of key performance indicators of the interface agent produced in the future, β1, β2, β3 and β4 are all regression coefficients, λ is the random error term, ∆γ represents the change value of surface tension, ∆η represents the change value of viscosity, ∆p represents the change value of pH value, ∆g represents the change value of drying time, and ∆f represents the change value of adhesion.

10. The intelligent production management system based on interface agent performance test data according to claim 9, characterized in that: The specific process of the production parameter optimization module optimizing the production parameters using the established prediction model is as follows: Acquire real-time interface agent performance test data and real-time production parameters, convert them into corresponding prediction matrix GXJ and input them into the interface agent key performance indicator change prediction model. The interface agent real-time key performance indicator change matrix GXB produced in the future is obtained through the model, and the obtained interface agent real-time key performance indicator change matrix GXB is sent to the production management platform; After receiving the real-time key performance indicator change matrix GXB of the interface agent, the production management platform immediately takes adjustment measures to optimize the production parameters in the interface agent production process.

Images