Equipment parameter cooperative control system and control method

Abstract

The invention discloses an equipment parameter cooperative control system and method, and belongs to the technical field of equipment parameter cooperative control, and the method comprises the steps: carrying out the multi-stage comparison and analysis of the deviation degree of a real-time production parameter value and an initial production parameter value, carrying out the sudden disturbance monitoring, and when the sudden disturbance of target equipment is detected, carrying out the control of the target equipment; performing reverse reasoning and positioning on abnormal equipment operation parameters of abnormal reasons causing sudden disturbance, generating regulation and control instructions of the abnormal equipment operation parameters, and performing dynamic updating on the regulation and control instructions of the abnormal equipment operation parameters based on the abnormal production parameter values; and after the production task is completed, carrying out quality detection on the product, and if the product quality detection is not qualified, carrying out product quality early warning and generating a product quality production report. The problems that target equipment depends on preset process parameters in production, the sudden disturbance in the production process lacks real-time dynamic compensation capacity, and the production quality is affected are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative control of equipment parameters, and more specifically to a collaborative control system and method for equipment parameters. Background Art

[0002] In modern industrial production, most traditional production equipment relies on preset process parameters during operation. However, the actual production environment is complex and changeable, with many uncontrollable factors. For example, during the production process, sudden disturbances such as fluctuations in the composition of raw materials, wear and aging of equipment components, and changes in environmental temperature and humidity often occur. When these situations occur, existing production equipment lacks the ability of real-time dynamic compensation and cannot adjust the operating parameters in time to adapt to these changes. As a result, during the production process, quality problems such as uneven internal structure, inconsistent density, and deviation in dimensional accuracy are likely to occur. At the same time, due to the inability to flexibly adjust according to the actual production situation, the production efficiency is also greatly limited, and the scrap rate remains high. Therefore, in order to overcome these limitations, the present invention proposes a collaborative control system and method for equipment parameters. Summary of the Invention

[0003] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a collaborative control system and method for equipment parameters, which solves the problem that production equipment relies on preset process parameters during production and lacks the ability of real-time dynamic compensation in the face of sudden disturbances during the production process, affecting production quality.

[0004] To achieve the above purpose, the present invention provides the following technical solutions:

[0005] A collaborative control system for equipment parameters includes a parameter acquisition module, an analysis and decision-making module, an equipment control module, and a quality detection module;

[0006] The parameter acquisition module is used to determine the initial production parameter values based on the production tasks of the target equipment, collect the real-time production parameter values during the execution of the production tasks by the target equipment, and perform multi-level comparative analysis on the real-time production parameters and their deviation degrees to monitor sudden disturbances and determine whether there are sudden disturbances in the target equipment;

[0007] The analysis and decision-making module is used to, when detecting that there are sudden disturbances in the target equipment, construct a parameter abnormal propagation path map through Bayes' theorem to locate the abnormal production parameters, and based on the abnormal production parameters, inversely infer and locate the abnormal equipment operating parameters that cause the sudden disturbances;

[0008] The equipment control module is used to sequentially generate control instructions for the abnormal equipment operating parameters according to the arrangement order of the abnormal equipment operating parameters, and dynamically update the control instructions for the abnormal equipment operating parameters based on the real-time deviation degree of the abnormal production parameter values;

[0009] The quality inspection module is used to collect quality parameter values after the target device completes the production task, perform quality inspection on the product, and judge whether to issue a product quality warning and generate a product quality production report according to the product quality inspection results.

[0010] Specifically, the specific steps of sudden disturbance monitoring include:

[0011] For each production parameter, extract the variance of the historical production parameters under the same production task from the historical production task database, and combine it with its initial production parameter value to set production parameter thresholds, including the upper production parameter threshold and the lower production parameter threshold;

[0012] Based on the production parameter thresholds, perform potential disturbance judgment on the collected real-time production parameter values, and compare each real-time production parameter value with the corresponding production parameter threshold in real time;

[0013] If the real-time production parameter value is greater than the upper production parameter threshold or less than the lower production parameter threshold, mark the real-time production parameter as an abnormal production parameter, mark the real-time production parameter value as an abnormal production parameter value, and record the occurrence time of the abnormal production parameter value; otherwise, do not perform any processing;

[0014] Configure the deviation degree threshold, and calculate the ratio of the absolute value of the difference between the abnormal production parameter value and the initial production parameter value to the initial production parameter value as the deviation degree of the abnormal production parameter value. If the deviation degree of the abnormal production parameter value is greater than the deviation degree threshold, issue a sensor abnormality warning; otherwise, start abnormal monitoring for the real-time production parameter marked as an abnormal production parameter.

[0015] Specifically, the specific steps of abnormal monitoring include:

[0016] Configure the abnormal duration threshold, obtain the production parameter abnormal duration marked as an abnormal production parameter. If the abnormal duration is greater than the abnormal duration threshold, perform fitting prediction on the production parameter values marked as abnormal production parameters through a time series fitting algorithm to obtain the production parameter prediction values within the future abnormal duration threshold;

[0017] Configure the mutation threshold, compare the production parameter prediction values with the corresponding production parameter thresholds in real time, and count the number of production parameter prediction values marked as abnormal production parameter values. If the number is greater than the mutation threshold, it is determined that there is a sudden disturbance; otherwise, continue to perform abnormal monitoring on the real-time production parameters marked as abnormal production parameters.

[0018] Specifically, the specific steps of obtaining the production parameter prediction values within the future abnormal duration threshold include:

[0019] Obtain the production parameter values marked as abnormal production parameters from the start time of the production task of the target device to the current moment, and arrange them in chronological order to construct an abnormal prediction set;

[0020] After preprocessing the data of the abnormal prediction set, divide it into a training set and a test set, select a time series fitting algorithm, construct an abnormal prediction model, and fit the abnormal prediction model through the training set and the test set;

[0021] According to the abnormal duration threshold, predict the predicted production parameter values within the future abnormal duration threshold time period through the fitted abnormal prediction model.

[0022] Specifically, the specific steps for constructing the parameter abnormal propagation path graph include:

[0023] According to the production task type, screen the historical production task data of the target device from the historical production task database to construct an abnormal propagation data set. The variables in the abnormal propagation data set include historical production parameter variables and historical device operation parameter variables;

[0024] Clean the data of the abnormal propagation data set, remove the noise data, duplicate data and error data in the abnormal propagation data set; and align the time stamps of the abnormal propagation data set according to the acquisition time of the variable values in the abnormal propagation data set;

[0025] Preliminarily set the preliminary causal relationship between each historical production parameter variable and each historical device operation parameter variable in the abnormal propagation data set, and according to the Granger causality test, by comparing the time sequence and correlation of different variables in the abnormal propagation data set, calculate the causal relationship strength and significance level between each parameter, set the significance level threshold, and screen the causal relationship parameter pairs with a significance level greater than the significance level threshold;

[0026] Define the abnormal situation of each production parameter as an abnormal event, and count the frequency of each production parameter marked as an abnormal production parameter according to the abnormal propagation data set as the prior probability of the production parameter appearing in an abnormal situation;

[0027] For the parameter pairs with causal relationships, according to Bayes' theorem, by calculating the prior probability of each parameter pair with causal relationships, calculate the conditional probability of another parameter being abnormal under the condition that one parameter is abnormal.

[0028] Specifically, the specific steps for constructing the parameter abnormal propagation path graph also include:

[0029] Construct a framework for the abnormal propagation path graph of production parameters, taking each production parameter as a node in the graph, where each node represents the abnormal state of a specific production parameter; add edges to connect parameter pairs with causal relationships, and the direction of the edge represents the direction of the causal relationship, pointing from the cause node to the result node; and use the conditional probability of the parameter pairs with causal relationships as the weight of the edge.

[0030] Specifically, the specific steps for locating the abnormal device operation parameters that cause the abnormal reasons for sudden disturbances are as follows:

[0031] Take the nodes of the parameter abnormal propagation path graph corresponding to the abnormal production parameters of the target device as the starting nodes;

[0032] Starting from the starting node, trace back along the edges in the parameter abnormal propagation path graph, and according to the weight of the edge, calculate the posterior probability of each cause node causing the abnormality of the starting node;

[0033] Configure a probability threshold, and based on the calculated posterior probability, screen out the cause nodes with a posterior probability greater than the probability threshold as potential abnormal causes;

[0034] Determine the abnormal device operation parameters according to the potential abnormal causes, and sort the abnormal device operation parameters in descending order according to the posterior probability of the potential abnormal causes.

[0035] Specifically, the specific steps for generating a regulation instruction for abnormal device operation parameters are as follows:

[0036] Obtain the arrangement order of the abnormal production parameters, and generate a regulation instruction for the abnormal device operation parameters according to the arrangement order, that is:

[0037] Screen the extreme values of the historical device operation parameter values of the current production task from the historical production task database according to the production task type to set the operating standard range of the abnormal device operation parameters;

[0038] Determine the preliminary regulation direction according to the deviation direction between the abnormal production parameter value and the initial production parameter value, including: if the abnormal production parameter value is greater than the initial production parameter value, then reduce the abnormal device operation parameter value, and if the abnormal production parameter value is less than the initial production parameter value, then increase the abnormal device operation parameter value;

[0039] Set a basic step size for each device operation parameter, divide the deviation degree of the abnormal production parameter value into segments, each segment of deviation degree matches a mapping coefficient, and use a segmented mapping function to determine the mapping coefficient according to the deviation degree between the current abnormal production parameter value and the initial production parameter value to regulate the step size length;

[0040] Obtain the abnormal device operation parameter values of the current target device, and generate a regulation instruction sequence according to the operating standard range, preliminary regulation direction, and post-regulation step length of the abnormal device operation parameters.

[0041] Specifically, the specific steps for dynamically updating the regulation instructions for abnormal device operation parameters include:

[0042] Execute the regulation instruction sequence of the abnormal device operation parameters in sequence according to the arrangement order of the abnormal device operation parameters;

[0043] Configure a stop threshold. For the regulation instruction sequence of any abnormal device operation parameter, after executing each regulation instruction in the regulation instruction sequence, monitor the deviation degree of the abnormal production parameter. If the deviation degree of the abnormal production parameter is greater than the stop threshold, then determine whether the deviation degree is alleviated;

[0044] When the deviation degree of the abnormal production parameter decreases or remains unchanged, it is determined that the deviation degree is alleviated, otherwise it is determined that it is not alleviated. If the deviation degree is alleviated, continue to execute the next regulation instruction in the regulation instruction sequence until the regulation instruction sequence of the current abnormal device operation parameter is executed completely;

[0045] If the deviation degree is not alleviated, after marking the current regulation instruction as a regulation instruction abnormal point, continue to execute the next regulation instruction in the regulation instruction sequence for regulation instruction monitoring, configure an abnormal point threshold, count the number of regulation instruction abnormal points. If it is greater than the abnormal point threshold, update the regulation direction to the opposite direction of the preliminary regulation direction, and update it in combination with the current abnormal device operation parameter value according to the operating standard range and post-regulation step length of the abnormal device operation parameters to obtain an updated regulation instruction sequence, and perform abnormal monitoring on the updated regulation instruction sequence; otherwise continue to execute the next regulation instruction in the regulation instruction sequence until the regulation instruction sequence of the current abnormal device operation parameter is executed completely;

[0046] If the deviation degree of the abnormal production parameter is less than or equal to the stop threshold, stop the regulation of the device operation parameters.

[0047] Specifically, the specific steps for performing abnormal monitoring on the updated regulation instruction sequence include:

[0048] After executing the regulation instruction of the updated regulation instruction sequence, obtain the deviation degree of the abnormal production parameter. If the deviation degree is alleviated, continue to execute the next regulation instruction in the updated regulation instruction sequence until the updated regulation instruction sequence of the current abnormal production parameter is executed completely;

[0049] If the deviation degree is not alleviated, mark the abnormal points of the updated control instruction, conduct monitoring of the updated control instruction, count the number of abnormal points of the updated control instruction. If it is greater than the abnormal point threshold, stop the control and issue a control abnormality warning. Otherwise, continue to execute the next control instruction in the updated control instruction sequence until the updated control instruction sequence of the current abnormal production parameter is executed completely;

[0050] When the control instruction sequences of all the operation parameters of the malfunctioning devices are executed, if the deviation degree of the abnormal production parameter is greater than the stop threshold, issue a control abnormality warning.

[0051] A method for collaborative control of device parameters, comprising the following steps:

[0052] Step S1: Determine the initial production parameter values based on the production tasks of the target device, collect the real-time production parameter values during the execution of the production tasks by the target device, and conduct sudden disturbance monitoring by performing multi-level comparative analysis on the real-time production parameters and their deviation degrees to determine whether there is a sudden disturbance in the target device;

[0053] Step S2: When it is detected that there is a sudden disturbance in the target device, construct a parameter abnormal propagation path map to locate the abnormal production parameter through Bayes' theorem, and based on the abnormal production parameter, inversely infer and locate the abnormal device operation parameter that causes the sudden disturbance;

[0054] Step S3: Generate control instructions for the abnormal device operation parameters in sequence according to the arrangement order of the abnormal device operation parameters that cause the sudden disturbance, and dynamically update the control instructions for the abnormal device operation parameters based on the real-time deviation degree of the abnormal production parameter value;

[0055] Step S4: After the target device completes the production task, collect the quality parameter values, conduct quality inspection on the product. If the product quality inspection is unqualified, issue a product quality warning and generate a product quality production report.

[0056] Advantages of the present invention:

[0057] Equipment parameter collaborative control system and control method, which effectively solve the problem that production equipment relies on preset process parameters during production, lacks real-time dynamic compensation ability in the face of sudden disturbances, and affects production quality. The parameter acquisition module analyzes real-time production parameters through multi-level comparison to accurately monitor sudden disturbances and promptly detect abnormal situations during the production process; the analysis and decision-making module uses Bayes' theorem to construct a graph for reverse reasoning to quickly locate abnormal equipment operation parameters that cause sudden disturbances; the equipment control module generates and dynamically updates control instructions based on these parameters to achieve real-time dynamic compensation for sudden disturbances, effectively reducing the impact of sudden disturbances on production. After production is completed, the quality detection module comprehensively detects the product quality, warns of unqualified products, and generates a report in combination with multi-module data, providing strong data support for optimizing the initial production parameter setting, adjusting the production parameter threshold, and improving the overall production process in subsequent production, thereby comprehensively improving product quality and production efficiency and ensuring production stability and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flowchart of the specific steps for monitoring sudden disturbances of the present invention;

[0059] Figure 2 It is a flowchart of the specific steps for abnormal monitoring of the present invention;

[0060] Figure 3 It is a flowchart of the specific steps for dynamically updating the control instructions for abnormal equipment operation parameters of the present invention;

[0061] Figure 4 It is a flowchart of the specific steps for abnormal monitoring of the updated control instruction sequence of the present invention;

[0062] Figure 5 It is a flowchart of a method for collaborative control of equipment parameters of the present invention;

[0063] Figure 6 It is a schematic structural diagram of a collaborative control system for equipment parameters of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] Example 1

[0065] Please refer to Figure 1 , this embodiment introduces a collaborative control system for equipment parameters, including: a parameter acquisition module, an analysis and decision-making module, an equipment control module, and a quality detection module;

[0066] The parameter acquisition module is used to determine the initial production parameter values based on the production tasks of the target equipment, collect real-time production parameter values during the execution of the production tasks by the target equipment, and perform multi-level comparison and analysis on the real-time production parameters and their deviation degrees to monitor sudden disturbances and determine whether there are sudden disturbances in the target equipment;

[0067] In this embodiment, the target device refers to a vertical roll centrifuge and its key execution units, including a drum system, a mold device, a molten metal pouring system, etc. According to the current production task of the vertical roll centrifuge, the initial production parameters matching the current production task are determined based on the historical production task database. The production parameters include drum speed, mold preheating temperature, molten metal pouring flow rate, and molten metal temperature. During the execution of the production task by the vertical roll centrifuge, the parameter acquisition module constructs an all-round data acquisition network through multi-type and high-precision sensors. At key parts of the vertical roll centrifuge, including the drum, feed pipe, cooling system, etc., sensors such as speed sensors, temperature sensors, pressure sensors, and vibration sensors are respectively deployed. The real-time production parameters are collected in a high-frequency and uninterrupted manner to ensure the timeliness and integrity of the data. The real-time production parameters are compared with the dynamic threshold range set according to historical data and process standards. Once the data exceeds the threshold range, it is preliminarily determined that there is a sudden disturbance. At the same time, the change trend and characteristics of the real-time production parameters are analyzed to further determine whether there is a sudden disturbance, providing a solid data basis for subsequent analysis and decision-making.

[0068] Please refer to Figure 2 , preferably, the specific steps of sudden disturbance monitoring include:

[0069] For each production parameter, extract the variance of the historical production parameters under the same production task from the historical production task database, and combine it with its initial production parameter value to set the production parameter threshold, including the production parameter upper threshold and the production parameter lower threshold; Exemplarily, according to the statistical principle of the 3σ criterion, set the production parameter threshold, add 3 times the standard deviation to the initial production parameter value as the production parameter upper threshold, and subtract 3 times the standard deviation from the initial production parameter value as the production parameter lower threshold. The historical production task database is a database system specifically used to store and manage various historical data of production parameters generated during the production process of the vertical roll centrifuge, including production parameter data, production task data, equipment operation data, quality inspection data, and personnel operation data;

[0070] Based on the production parameter threshold, perform a potential disturbance judgment on the collected real-time production parameter values. Compare each real-time production parameter value with the corresponding production parameter threshold in real time. If the real-time production parameter value is greater than the production parameter upper threshold or less than the production parameter lower threshold, mark the real-time production parameter as an abnormal production parameter, mark the real-time production parameter value as an abnormal production parameter value, and record the occurrence time of the abnormal production parameter value; otherwise, do not perform any processing; Comparing the collected production parameter values with the pre-set production parameter threshold in real time can timely capture the abnormal changes of parameters in the production process, provide detailed time series information for subsequent in-depth analysis and processing, facilitate tracing and troubleshooting potential production faults, and effectively reduce the losses caused by sudden disturbances to production.

[0071] According to the production process requirements and equipment characteristics, configure the deviation degree threshold. By calculating the ratio of the absolute value of the difference between the abnormal production parameter value and the initial production parameter value to the initial production parameter value as the deviation degree of the abnormal production parameter value. If the deviation degree of the abnormal production parameter value is greater than the deviation degree threshold, sensor abnormal early warning is carried out. Otherwise, abnormal monitoring is started for the real-time production parameters marked as abnormal production parameters. Configuring the deviation degree threshold and calculating the deviation degree of the abnormal production parameter value can further distinguish whether the parameter deviation is caused by sensor abnormality or a real potential disturbance in the production process. When the deviation degree exceeds the threshold, sensor abnormal early warning is carried out, which can timely detect and handle sensor failures and avoid misjudgment caused by sensor errors. And starting abnormal monitoring for abnormal production parameters that do not reach the sensor abnormal standard can continuously track potential problems in the production process to ensure that any possible sudden disturbance is not missed.

[0072] Please refer to Figure 3 , preferably, the specific steps of the abnormal monitoring include:

[0073] Configure the abnormal duration threshold, obtain the abnormal duration of the real-time production parameters marked as abnormal production parameters. If the abnormal duration is greater than the abnormal duration threshold, the production parameter values marked as abnormal production parameters are fitted and predicted through the time series fitting algorithm to obtain the production parameter prediction values within the future abnormal duration threshold. Otherwise, no processing is carried out. The time series fitting algorithm includes the ARIMA model or the LSTM neural network. Obtain the production task start time of the vertical roll centrifuge to the current production parameter values marked as abnormal production parameters to construct an abnormal prediction set. After preprocessing operations such as data cleaning and normalization on the abnormal prediction set, input it into the fitted abnormal prediction model to model and analyze the change trend of the abnormal production parameters and predict the production parameter prediction values within the future abnormal duration threshold period.

[0074] Preferably, the specific steps of predicting the production parameter prediction values within the future abnormal duration threshold period include:

[0075] Obtain the production parameter values marked as abnormal production parameters from the production task start time of the vertical roll centrifuge to the current moment and arrange them in chronological order to construct an abnormal prediction set;

[0076] After preprocessing the abnormal prediction set, divide it into a training set and a test set. Data preprocessing includes data cleaning and missing value processing;

[0077] Select a time series fitting algorithm to construct an anomaly prediction model, and fit the anomaly prediction model through the training set and the test set; if the data in the anomaly prediction set is stationary and mainly shows a linear trend, select the ARIMA model; if the data has complex non-linear characteristics and long-term dependence relationships, select the LSTM neural network.

[0078] According to the anomaly duration threshold, determine the length of the future time period to be predicted, that is, predict the predicted values of production parameters within the future anomaly duration threshold time period. Input the production parameter values before the current moment into the fitted anomaly prediction model to predict the future production parameter values, and obtain the predicted values of production parameters within the future anomaly duration threshold time period. Determining the length of the predicted future time period according to the anomaly duration threshold can make the prediction results closely related to the actual production situation and reasonably estimate the changes in future production parameters. The obtained predicted values of production parameters can provide forward-looking guidance for the adjustment and control of the production process.

[0079] Configure a mutation threshold, compare the predicted values of production parameters with the corresponding production parameter thresholds in real time, and count the number of predicted values of production parameters marked as abnormal production parameter values. If this number is greater than the mutation threshold, it is determined that there is a sudden disturbance; otherwise, continue to monitor the real-time production parameters marked as abnormal production parameters for anomalies; counting the number of predicted values of production parameters marked as abnormal production parameter values and comparing it with the mutation threshold provides a quantitative judgment criterion, making the judgment of sudden disturbances more objective and accurate. When it is determined that there is a sudden disturbance when exceeding the mutation threshold, an alarm can be issued in a timely manner to prompt the enterprise to take corresponding measures for handling; otherwise, continue with anomaly monitoring to ensure continuous attention to the production process and not miss any potential risks.

[0080] The analysis and decision-making module is used to, when a sudden disturbance is detected in the vertical roll centrifuge, construct a parameter anomaly propagation path map through Bayes' theorem to locate abnormal production parameters, and based on the abnormal production parameters, inversely infer and locate the abnormal equipment operating parameters that cause the sudden disturbance.

[0081] Preferably, the specific steps for constructing the parameter anomaly propagation path map include:

[0082] According to the type of production tasks, historical production task data of vertical roll centrifuges is screened from the historical production task database to construct an abnormal propagation dataset. The variables in the abnormal propagation dataset include historical production parameter variables and historical equipment operation parameter variables; historical production parameter variables include drum speed, mold preheating temperature, molten metal pouring flow rate, and molten metal temperature; historical equipment operation parameter variables include the voltage, current, and frequency of the motor, mold heating power and heating time, molten metal pouring port size, pouring pressure and pouring time, molten metal furnace power and heat preservation time; the screened data is integrated to construct an abnormal propagation dataset. To ensure the integrity and accuracy of the dataset, each data item is recorded and annotated in detail, including information such as data source, collection time, and data meaning.

[0083] Data cleaning is performed on the abnormal propagation dataset to remove noise data, duplicate data, and error data in the abnormal propagation dataset; and the abnormal propagation dataset is timestamp-aligned according to the collection time of each variable value in the abnormal propagation dataset; a combination of multiple filtering algorithms is used, including median filtering, Kalman filtering, etc., to perform targeted processing on different types of data. For high-frequency fluctuating data, such as equipment vibration signals, Kalman filtering can effectively remove noise interference and retain the true signal characteristics; for low-frequency changing data, such as ambient temperature, median filtering is used to smooth the data and reduce the influence of random noise. Key information recorded in the abnormal propagation dataset, such as timestamps, parameter values, data sources, etc., is compared to accurately identify and delete duplicate data. For obviously incorrect data, such as parameter values outside the physical range, they are corrected according to business logic and historical data. If they cannot be corrected, they are marked as invalid data and excluded from the dataset. For parameter values with different collection frequencies, for example, production parameters are collected every 1 second, and equipment operation parameters are collected every 5 seconds, the linear interpolation method is used to interpolate the low-frequency data to high-frequency time points to ensure the consistency of all data in the time dimension.

[0084] Based on empirical knowledge, initially set the preliminary causal relationships between various historical production parameter variables and various historical equipment operation parameter variables in the abnormal propagation dataset. According to the experience and professional knowledge of experts in related fields, analyze the causal relationships existing between each parameter. For example, the change in the rotary drum speed will affect the distribution and forming effect of the molten metal, thus being related to the product quality; the level of the mold preheating temperature will affect the cooling rate of the molten metal, and further affect the microstructure of the product, etc. Take these causal relationships based on domain knowledge as the initial judgment basis. And according to the Granger causality test, by comparing the time sequence and correlation of different variables, judge whether one variable is the Granger cause of another variable, further verify the causal relationships between variables, calculate the causal relationship strength and significance level between each parameter, set the significance level threshold, and screen out the causal relationship parameter pairs with a significance level greater than the significance level threshold; for the parameter pairs with weak causal relationship strength and low significance level, eliminate them to reduce unnecessary interference and computational complexity.

[0085] Define the abnormal situation of each production parameter as an abnormal event. According to the abnormal propagation dataset, count the frequency of each production parameter being marked as an abnormal production parameter, which is used as the prior probability of the production parameter having an abnormal situation; according to the abnormal production parameter determination standard, define the abnormal event, conduct a comprehensive statistical analysis of each production parameter in the abnormal propagation dataset, and record the number of times each production parameter has an abnormality and the total number of data samples. By calculating the ratio of the number of abnormalities to the total number of samples, obtain the prior probability of each production parameter having an abnormal situation.

[0086] For the parameter pairs with causal relationships, according to Bayes' theorem, by calculating the prior probability of each parameter pair with causal relationships, calculate the conditional probability of another parameter being abnormal under the condition that one parameter is abnormal;

[0087] Construct the framework of the parameter abnormal propagation path graph. Take each production parameter as a node in the graph, and each node represents the abnormal state of a specific production parameter; add edges to connect the parameter pairs with causal relationships, and the direction of the edge represents the direction of the causal relationship, pointing from the cause node to the result node; and take the conditional probability of the parameter pair with causal relationships as the weight of the edge, indicating the possibility of the result node being abnormal under the condition that the cause node is abnormal.

[0088] Preferably, the specific steps for locating the abnormal equipment operation parameter that causes the sudden disturbance include:

[0089] Take the nodes of the parameter anomaly propagation path map corresponding to the abnormal production parameters of the vertical roll centrifuge as the starting nodes; this starting node represents the currently abnormal production parameter and is the starting point for subsequent backward reasoning and analysis. By clarifying the starting node and focusing on the causal relationship network related to the current abnormal production parameter, the pertinence and accuracy of the analysis are improved.

[0090] Starting from the starting node, trace back along the edges in the parameter anomaly propagation path map, and according to the weights of the edges, calculate the posterior probability of each possible cause node leading to the anomaly of the starting node.

[0091] Configure a probability threshold, and based on the calculated posterior probability, screen out the cause nodes with posterior probability greater than the probability threshold as potential abnormal causes; by screening out the cause nodes with posterior probability greater than the probability threshold, attention can be focused on those causes that are most likely to lead to the anomaly of the starting node, reducing the unnecessary analysis and troubleshooting scope and improving the efficiency of problem-solving.

[0092] Determine the abnormal equipment operation parameters according to the potential abnormal causes, and sort the abnormal equipment operation parameters in descending order according to the posterior probability of the potential abnormal causes.

[0093] The equipment control module is used to sequentially generate control instructions for the abnormal equipment operation parameters according to the arrangement order of the abnormal equipment operation parameters of the abnormal causes leading to the sudden disturbance, and dynamically update the control instructions for the abnormal equipment operation parameters based on the real-time deviation degree of the abnormal production parameter values;

[0094] Preferably, the specific steps for generating the control instructions for the abnormal equipment operation parameters include:

[0095] Obtain the arrangement order of the abnormal production parameters, and generate control instructions for the abnormal equipment operation parameters according to the arrangement order;

[0096] Preferably, the specific steps for the control instructions of the abnormal production parameters include:

[0097] Screen the extreme values of the historical equipment operation parameter values of the current production task from the historical production task database according to the production task type to set the operating standard range of the abnormal equipment operation parameters;

[0098] Determine the preliminary control direction according to the deviation direction between the abnormal production parameter value and the initial production parameter value, including: if the abnormal production parameter value is greater than the initial production parameter value, then reduce the abnormal equipment operation parameter value, and if the abnormal production parameter value is less than the initial production parameter value, then increase the abnormal equipment operation parameter value;

[0099] Set a basic step size for each device operating parameter, segmentally divide the deviation degree of abnormal production parameter values, match a mapping coefficient for each segment of the deviation degree, and adopt a segmented mapping function to determine the mapping coefficient according to the deviation degree between the current abnormal production parameter value and the initial production parameter value to regulate the step size length. For example, divide the deviation degree into multiple intervals, including a mild deviation interval, a moderate deviation interval, and a severe deviation interval. Match a mapping coefficient for each interval. For example, the mapping coefficient for the mild deviation interval is 0.5, the mapping coefficient for the moderate deviation interval is 1, and the mapping coefficient for the severe deviation interval is 2. Determine the interval to which it belongs based on the calculated deviation degree, and then multiply the basic step size by the corresponding mapping coefficient to obtain the regulated step size length.

[0100] Obtain the current abnormal device operating parameter value of the vertical roll centrifuge as the initial value, and generate a regulation instruction sequence according to the operating standard range, preliminary regulation direction, and regulated step size length of the abnormal device operating parameter.

[0101] Please refer to Figure 4 , preferably, the specific steps for dynamically updating the regulation instructions for abnormal device operating parameters include:

[0102] Execute the regulation instruction sequence of the abnormal device operating parameter sequentially according to the arrangement order of the abnormal device operating parameters; this arrangement order is usually arranged from high to low according to the posterior probability of potential abnormal causes, and first execute the regulation instruction sequence of the device operating parameter with a greater impact on the abnormality.

[0103] Configure a stop threshold. For the regulation instruction sequence of any abnormal device operating parameter, after executing each regulation instruction in the regulation instruction sequence, monitor the deviation degree of the abnormal production parameter. If the deviation degree of the abnormal production parameter is greater than the stop threshold, then determine whether the deviation degree is alleviated; and timely obtain the change situation of the abnormal production parameter during the device operation process, provide a data basis for the regulation effect evaluation, accurately judge whether the regulation is effective, so as to decide the subsequent regulation strategy.

[0104] When the deviation degree of the abnormal production parameter decreases or remains unchanged, it is determined that the deviation degree is alleviated, otherwise it is determined that it is not alleviated. If the deviation degree is alleviated, continue to execute the next regulation instruction in the regulation instruction sequence until the regulation instruction sequence of the current abnormal device operating parameter is executed; for effective regulation, continuously promote the regulation process and gradually restore the device operating parameter to normal.

[0105] If the degree of deviation is not alleviated, the current control instruction is marked as a control instruction abnormal point, and the next control instruction in the control instruction sequence is executed. If the degree of deviation is still not alleviated, the next control instruction is marked as a control instruction abnormal point. During the execution of the control instruction sequence, the control instruction is monitored, the abnormal point threshold is configured, and the number of control instruction abnormal points is counted. If the number of abnormal points in the control instruction sequence is greater than the abnormal point threshold, the control direction is updated to the opposite direction of the initial control direction. Combined with the current abnormal equipment operating parameter value, the operating standard range of the abnormal equipment operating parameter and the step length after control are updated to obtain an updated control instruction sequence, and the updated control instruction sequence is monitored for abnormalities. Otherwise, the next control instruction in the control instruction sequence is executed until the control instruction sequence of the current abnormal equipment operating parameter is completed. For invalid control, the control strategy is adjusted in time to avoid the continuation of invalid operations and improve the pertinence and effectiveness of control.

[0106] If the deviation degree of the abnormal production parameters is less than or equal to the stop threshold, the equipment operation parameter control is stopped;

[0107] See also Figure 5 Preferably, the specific steps of performing abnormal monitoring on the update control instruction sequence include:

[0108] After executing the control instruction of the updated control instruction sequence, obtain the degree of deviation of the abnormal production parameter. If the deviation degree is alleviated, continue to execute the next control instruction of the updated control instruction sequence until the updated control instruction sequence of the current abnormal production parameter is completed; continuously track the effect of the updated control instruction to ensure that the new control strategy can effectively improve the equipment operation status.

[0109] If the degree of deviation is not alleviated, the abnormal points of the updated control instructions will be marked, and the updated control instructions will be monitored and the number of abnormal points of the updated control instructions will be counted. If it is greater than the abnormal point threshold, the control will be stopped and an abnormal control warning will be issued. Otherwise, the next control instruction in the updated control instruction sequence will continue to be executed until the updated control instruction sequence of the current abnormal production parameters is completed. If the updated control instruction is still invalid, it will be discovered in time to attract the attention of relevant personnel through the early warning mechanism so that further measures can be taken.

[0110] When all control instruction sequences for abnormal equipment operating parameters have completed execution, if the deviation of abnormal production parameters exceeds the stop threshold, a control anomaly warning is issued. A comprehensive assessment of the final results of the entire control process ensures that equipment operating parameters are stable and meet production requirements. If control targets are not met, timely warnings are issued, providing a basis for subsequent in-depth analysis and resolution, ensuring smooth production.

[0111] The quality inspection module is used to collect quality parameter values after the vertical roll centrifuge completes the production task, conduct quality inspection on the roll products. If the quality inspection of the roll products fails, product quality early warning is carried out, and combined with the real-time production parameter values in the production process obtained by the parameter collection module, the abnormal equipment operation parameters for the analysis and decision-making module to locate the abnormal causes of sudden disturbances, and the records of the regulation instructions of the equipment control module for the abnormal equipment operation parameters, a product quality production report is generated. It provides data support for the initial production parameter setting, production parameter threshold adjustment and overall production process optimization of subsequent production tasks, so as to continuously improve product quality and production efficiency.

[0112] Preferably, the specific steps for conducting quality inspection on the roll products include:

[0113] After the vertical roll centrifuge completes the production task, start the quality inspection process, collect quality parameter values for the key quality characteristics of the roll products, including: measuring the surface roughness of the roll using a profilometer, obtaining the dimensional accuracy of the roll, including diameter, length, cylindricity, through a coordinate measuring machine; detecting the hardness values at different positions on the surface and inside of the roll using a hardness tester; performing non-destructive testing using an ultrasonic flaw detector and a magnetic particle flaw detector to check for defects such as cracks and sand holes inside and on the surface.

[0114] Record the various quality parameter values collected in real time, establish a quality parameter database, and associate each quality parameter with its corresponding product batch, production time, and equipment number to ensure the accuracy and integrity of the data.

[0115] According to the quality standards of the roll products, set the quality parameter range, compare the collected quality parameter values with the corresponding quality parameter range one by one. If all quality parameters meet the quality standards, it is determined that the quality of the roll product is qualified, the current quality inspection process is ended, and the product is marked as a qualified product and enters the next link; if any one or more quality parameters exceed the standard range, it is determined that the quality of the roll product is unqualified, start the quality problem analysis process, and generate a product quality production report.

[0116] Preferably, the specific steps for generating a product quality production report include:

[0117] When the quality inspection of the roll product fails, connect with the parameter collection module to obtain the real-time production parameter values and abnormal production parameters during the production process of the roll product, and whether there are sudden disturbances.

[0118] Through the analysis and decision-making module, locate the abnormal equipment operation parameters that cause sudden disturbances; for example, whether the voltage, current, and frequency of the motor are stable, whether the mold heating power and heating time meet the requirements, whether the dimensions, pouring pressure, and pouring time of the molten metal pouring port are accurate, whether the power and heat preservation time of the molten metal furnace are appropriate, etc., and determine the impact mechanism of equipment operation anomalies on product quality.

[0119] Retrieve the control instruction records for abnormal equipment operation parameters from the equipment control module, including the control instructions taken for abnormal equipment operation parameters during the production process, as well as the execution time and execution results of the control instructions;

[0120] According to the information obtained above, generate a product quality production report. The report content includes product basic information, which includes batch number, serial number, production time, quality inspection results, review of production process parameters, analysis of abnormal equipment operation parameters, and control instructions; summarize and organize the generated product quality production report and related quality data, and store them in a long-term quality database. Provide historical references for setting initial production parameters for subsequent production tasks. By analyzing the relationship between product quality and initial production parameters of different batches, optimize the values of the initial parameters to make them more in line with the requirements of high-quality production. Based on a large amount of quality data and production process data, dynamically adjust the production parameter thresholds. According to the frequency and type of quality problems, reasonably expand or shrink the threshold ranges of certain parameters to improve the stability of the production process and the consistency of product quality.

[0121] Embodiment 2

[0122] Please refer to Figure 6 , this embodiment introduces a method for collaborative control of equipment parameters, including the following steps:

[0123] Step S1: Determine the initial production parameter values based on the production tasks of the target equipment, and collect real-time production parameter values during the execution of the production tasks by the target equipment. Through multi-level comparative analysis of the real-time production parameters and their deviation degrees, conduct sudden disturbance monitoring to determine whether there are sudden disturbances in the target equipment;

[0124] Step S2: When it is detected that there are sudden disturbances in the target equipment, through Bayes' theorem, construct a parameter abnormal propagation path map to locate the abnormal production parameters, and based on the abnormal production parameters, inversely infer and locate the abnormal equipment operation parameters that cause the sudden disturbances;

[0125] Step S3: According to the arrangement order of the abnormal equipment operation parameters that cause the sudden disturbances, sequentially generate control instructions for the abnormal equipment operation parameters, and dynamically update the control instructions for the abnormal equipment operation parameters based on the real-time deviation degree of the abnormal production parameter values;

[0126] Step S4: After the target device completes the production task, collect the quality parameter values, conduct quality inspection on the product. If the product quality inspection is unqualified, issue a product quality warning, and generate a product quality production report by combining the real-time production parameter values during the production process, the abnormal equipment operation parameters of the abnormal reasons causing the sudden disturbance, and the records of the regulation instructions for the abnormal equipment operation parameters.

[0127] Preferably, the specific steps of sudden disturbance monitoring include:

[0128] For each production parameter, extract the variance of the historical production parameters under the same production task from the historical production task database, and combine its initial production parameter value to set the production parameter threshold, including the upper production parameter threshold and the lower production parameter threshold;

[0129] Based on the production parameter threshold, conduct a potential disturbance judgment on the collected real-time production parameter values, and compare each real-time production parameter value with the corresponding production parameter threshold in real time;

[0130] If the real-time production parameter value is greater than the upper production parameter threshold or less than the lower production parameter threshold, mark the real-time production parameter as an abnormal production parameter, mark the real-time production parameter value as an abnormal production parameter value, and record the occurrence time of the abnormal production parameter value; otherwise, do not perform any processing;

[0131] Configure the deviation degree threshold. By calculating the ratio of the absolute value of the difference between the abnormal production parameter value and the initial production parameter value to the initial production parameter value as the deviation degree of the abnormal production parameter value, if the deviation degree of the abnormal production parameter value is greater than the deviation degree threshold, issue a sensor abnormality warning, otherwise start abnormal monitoring for the real-time production parameter marked as an abnormal production parameter.

[0132] Preferably, the specific steps of abnormal monitoring include:

[0133] Configure the abnormal duration threshold, obtain the production parameter abnormal duration marked as an abnormal production parameter. If the abnormal duration is greater than the abnormal duration threshold, perform fitting prediction on the production parameter values marked as an abnormal production parameter through the time series fitting algorithm to obtain the production parameter prediction values within the future abnormal duration threshold;

[0134] Configure the mutation threshold, compare the production parameter prediction values with the corresponding production parameter thresholds in real time, and count the number of production parameter prediction values marked as abnormal production parameter values. If the number is greater than the mutation threshold, it is determined that there is a sudden disturbance, otherwise continue to conduct abnormal monitoring on the real-time production parameter marked as an abnormal production parameter.

[0135] Preferably, the specific steps of obtaining the production parameter prediction values within the future abnormal duration threshold include:

[0136] Obtain the production parameter values marked as abnormal production parameters from the start time of the production task of the target device to the current moment, and arrange them in chronological order to construct an abnormal prediction set;

[0137] After preprocessing the data of the abnormal prediction set, divide it into a training set and a test set, select a time series fitting algorithm, construct an abnormal prediction model, and fit the abnormal prediction model through the training set and the test set;

[0138] According to the abnormal duration threshold, predict the predicted production parameter values within the future abnormal duration threshold period through the fitted abnormal prediction model.

[0139] Preferably, the specific steps for locating the abnormal device operation parameters that cause the abnormal reason of the sudden disturbance include:

[0140] Take the nodes of the parameter abnormal propagation path map corresponding to the abnormal production parameters of the target device as the starting nodes;

[0141] Starting from the starting node, trace back in reverse along the edges in the parameter abnormal propagation path map, and calculate the posterior probability of each cause node causing the abnormality of the starting node according to the weight of the edge;

[0142] Configure a probability threshold, and according to the calculated posterior probability, screen out the cause nodes with a posterior probability greater than the probability threshold as potential abnormal reasons;

[0143] Determine the abnormal device operation parameters according to the potential abnormal reasons, and sort the abnormal device operation parameters in descending order according to the posterior probability of the potential abnormal reasons.

[0144] Preferably, the specific steps for generating a regulation instruction for the abnormal device operation parameters include:

[0145] Obtain the arrangement order of the abnormal production parameters, and generate a regulation instruction for the abnormal device operation parameters according to the arrangement order, that is:

[0146] Screen the extreme values of the historical device operation parameter values of the current production task from the historical production task database according to the production task type to set the operating standard range of the abnormal device operation parameters;

[0147] Determine the preliminary regulation direction according to the deviation direction between the abnormal production parameter value and the initial production parameter value, including: if the abnormal production parameter value is greater than the initial production parameter value, then reduce the abnormal device operation parameter value, and if the abnormal production parameter value is less than the initial production parameter value, then increase the abnormal device operation parameter value;

[0148] Set a basic step size for each device operation parameter, segment the deviation degree of abnormal production parameter values, match a mapping coefficient for each segment of the deviation degree, adopt a segmented mapping function, and determine the mapping coefficient according to the deviation degree between the current abnormal production parameter value and the initial production parameter value to regulate the step size length;

[0149] Obtain the abnormal device operation parameter value of the current target device, and generate a regulation instruction sequence according to the operation standard range, preliminary regulation direction, and regulated step size length of the abnormal device operation parameter.

[0150] Working principle and its effects:

[0151] For the device parameter collaborative control system and control method of the present invention, the parameter acquisition module determines the initial production parameter value according to the production task and sets a threshold value. During production, by collecting real-time production parameters, it conducts multi-level comparison and analysis with the threshold value and the deviation degree to achieve sudden disturbance monitoring, can timely detect abnormal situations where the production deviates from the preset parameters, and solves the problem of relying on preset process parameters in the past and being unable to detect sudden disturbances in a timely manner. Once a sudden disturbance is detected, the analysis and decision-making module uses Bayes' theorem to screen data from the historical production task database to construct an abnormal propagation data set, and constructs a parameter abnormal propagation path map through steps such as cleaning and causal relationship analysis to locate the abnormal production parameters, and inversely infers and locates the abnormal device operation parameters that cause the sudden disturbance, providing a basis for subsequent precise regulation, and changing the situation of being difficult to locate the root cause of the problem in the face of sudden disturbances in the past. The device control module generates regulation instructions according to the arranged order of the located abnormal device operation parameters, and dynamically updates according to the real-time deviation degree of the abnormal production parameter value, realizing real-time dynamic compensation for sudden disturbances, effectively improving the ability to cope with sudden disturbances, and ensuring stable production. After production is completed, the quality detection module collects quality parameter values through various detection means to detect the product quality. If it is unqualified, it generates a report in combination with the data of other modules, providing data support for optimizing the initial production parameter setting, adjusting the production parameter threshold, and improving the overall production process in subsequent production, continuously improving the product quality and production efficiency, and solving the problem that the production quality is affected due to the lack of real-time dynamic compensation ability.

[0152] The above is only the preferred implementation mode of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A device parameter collaborative control system, characterized in that, It includes a parameter acquisition module, an analysis and decision-making module, a device control module, and a quality detection module; The parameter acquisition module is used to determine the initial production parameter values based on the production tasks of the target device, collect the real-time production parameter values during the execution of the production tasks by the target device, and perform multi-level comparative analysis on the deviation degree between the real-time production parameters to conduct sudden disturbance monitoring to determine whether there is a sudden disturbance in the target device; The analysis and decision-making module is used to, when detecting that there is a sudden disturbance in the target device, construct a parameter abnormal propagation path map through Bayes' theorem to locate the abnormal production parameters, and based on the abnormal production parameters, inversely infer and locate the abnormal device operation parameters that cause the sudden disturbance; The device control module is used to sequentially generate the regulation instructions for the abnormal device operation parameters according to the arrangement order of the abnormal device operation parameters, and dynamically update the regulation instructions for the abnormal device operation parameters based on the real-time deviation degree of the abnormal production parameter values; The quality detection module is used to collect the quality parameter values after the target device completes the production tasks, conduct quality detection on the products, and judge whether to issue a product quality early warning and generate a product quality production report according to the product quality detection results.

2. The device parameter collaborative control system according to claim 1, characterized in that, The specific steps of the sudden disturbance monitoring include: For each production parameter, extract the variance of the historical production parameters under the same production task from the historical production task database, and combine it with its initial production parameter value to set the production parameter threshold, including the production parameter upper threshold and the production parameter lower threshold; Based on the production parameter threshold, conduct potential disturbance judgment on the collected real-time production parameter values, and compare each real-time production parameter value with the corresponding production parameter threshold in real time; If the real-time production parameter value is greater than the production parameter upper threshold or less than the production parameter lower threshold, mark the real-time production parameter as an abnormal production parameter, mark the real-time production parameter value as an abnormal production parameter value, and record the occurrence time of the abnormal production parameter value; otherwise, do not perform any processing; Configure the deviation degree threshold, and calculate the ratio of the absolute value of the difference between the abnormal production parameter value and the initial production parameter value to the initial production parameter value as the deviation degree of the abnormal production parameter value. If the deviation degree of the abnormal production parameter value is greater than the deviation degree threshold, issue a sensor abnormal early warning; otherwise, start abnormal monitoring on the real-time production parameters marked as abnormal production parameters.

3. The device parameter collaborative control system according to claim 2, wherein The specific steps of the abnormal monitoring include: Configure the abnormal duration threshold, obtain the abnormal duration of the production parameter marked as an abnormal production parameter. If the abnormal duration is greater than the abnormal duration threshold, perform fitting prediction on the production parameter values marked as abnormal production parameters through the time series fitting algorithm to obtain the production parameter prediction values within the future abnormal duration threshold; Configure the mutation threshold, compare the production parameter prediction values with the corresponding production parameter thresholds in real time, and count the number of production parameter prediction values marked as abnormal production parameter values. If the number is greater than the mutation threshold, it is determined that there is a sudden disturbance; otherwise, continue to conduct abnormal monitoring on the real-time production parameters marked as abnormal production parameters. The specific steps for obtaining the predicted production parameter values within the future abnormal duration threshold are as follows: Obtain the production parameter values marked as abnormal production parameters from the start time of the production task of the target device to the current moment, and arrange them in chronological order to construct an abnormal prediction set; After preprocessing the data of the abnormal prediction set, divide it into a training set and a test set, select a time series fitting algorithm, construct an abnormal prediction model, and fit the abnormal prediction model through the training set and the test set; According to the abnormal duration threshold, predict the predicted production parameter values within the future abnormal duration threshold period through the fitted abnormal prediction model.

4. The device parameter collaborative control system according to claim 1, characterized in that The specific steps for constructing the parameter abnormal propagation path graph are as follows: According to the production task type, screen the historical production task data of the target device from the historical production task database to construct an abnormal propagation data set. The variables in the abnormal propagation data set include historical production parameter variables and historical device operation parameter variables; Clean the data of the abnormal propagation data set, remove the noise data, duplicate data, and error data in the abnormal propagation data set; and align the time stamps of the abnormal propagation data set according to the collection time of each variable value in the abnormal propagation data set; Preliminarily set the preliminary causal relationship between each historical production parameter variable and each historical device operation parameter variable in the abnormal propagation data set, and according to the Granger causality test, by comparing the chronological order and correlation of different variables in the abnormal propagation data set, calculate the causal relationship strength and significance level between each parameter, set the significance level threshold, and screen out the causal relationship parameter pairs with a significance level greater than the significance level threshold; Define the abnormal situation of each production parameter as an abnormal event, and count the frequency of each production parameter being marked as an abnormal production parameter according to the abnormal propagation data set as the prior probability of the production parameter appearing in an abnormal situation; For the parameter pairs with causal relationships, according to Bayes' theorem, by calculating the prior probability of each parameter pair with causal relationships, calculate the conditional probability of another parameter being abnormal under the condition that one parameter is abnormal.

5. The device parameter collaborative control system according to claim 4, characterized in that, The specific steps for constructing the parameter abnormal propagation path graph also include: Construct a framework for the parameter abnormal propagation path graph, take each production parameter as a node in the graph, and each node represents the abnormal state of a specific production parameter; Add edges to connect the parameter pairs with causal relationships, and the direction of the edge represents the direction of the causal relationship, pointing from the cause node to the result node; Take the conditional probability of the parameter pairs with causal relationships as the weight of the edge.

6. The device parameter collaborative control system according to claim 1, wherein The specific steps for locating the abnormal device operation parameters that cause the abnormal cause of the sudden disturbance are as follows: Take the nodes of the parameter abnormal propagation path graph corresponding to the abnormal production parameters of the target device as the starting nodes; Starting from the starting node, trace back along the edges in the parameter abnormal propagation path graph, and according to the weight of the edge, calculate the posterior probability of each cause node causing the abnormality of the starting node; Configure a probability threshold, and according to the calculated posterior probability, screen out the cause nodes with a posterior probability greater than the probability threshold as potential abnormal causes; Determine the abnormal device operation parameters according to the potential abnormal causes, and sort the abnormal device operation parameters in descending order of the posterior probability of the potential abnormal causes.

7. The device parameter collaborative control system according to claim 1, wherein The specific steps for generating the regulation instructions for the abnormal device operation parameters include: Obtain the arrangement order of the abnormal production parameters, and generate the regulation instructions for the abnormal device operation parameters according to the arrangement order, that is: Screen the extreme values of the historical device operation parameter values of the current production task from the historical production task database according to the production task type, so as to set the operation standard range of the abnormal device operation parameters; Determine the preliminary regulation direction according to the deviation direction between the abnormal production parameter value and the initial production parameter value, including: if the abnormal production parameter value is greater than the initial production parameter value, then reduce the abnormal device operation parameter value, and if the abnormal production parameter value is less than the initial production parameter value, then increase the abnormal device operation parameter value; Set a basic step size for each device operation parameter, divide the deviation degree of the abnormal production parameter value into segments, match a mapping coefficient for each segment of the deviation degree, and use a piecewise mapping function to determine the mapping coefficient according to the deviation degree between the current abnormal production parameter value and the initial production parameter value, so as to regulate the step size length; Obtain the abnormal device operation parameter value of the current target device, and generate a regulation instruction sequence according to the operation standard range, preliminary regulation direction and regulated step size length of the abnormal device operation parameters.

8. The device parameter collaborative control system according to claim 7, characterized in that, The specific steps for dynamically updating the regulation instructions for the abnormal device operation parameters include: Execute the regulation instruction sequence of the abnormal device operation parameters in turn according to the arrangement order of the abnormal device operation parameters; Configure a stop threshold. For the regulation instruction sequence of any abnormal device operation parameter, after executing each regulation instruction in the regulation instruction sequence, monitor the deviation degree of the abnormal production parameter. If the deviation degree of the abnormal production parameter is greater than the stop threshold, then judge whether the deviation degree is alleviated; When the deviation degree of the abnormal production parameter decreases or remains unchanged, it is determined that the deviation degree is alleviated, otherwise it is determined that it is not alleviated. If the deviation degree is alleviated, continue to execute the next regulation instruction in the regulation instruction sequence until the regulation instruction sequence of the current abnormal device operation parameter is executed; If the deviation degree is not alleviated, mark the current regulation instruction as an abnormal point of the regulation instruction, then continue to execute the next regulation instruction in the regulation instruction sequence, perform regulation instruction monitoring, configure an abnormal point threshold, count the number of abnormal points of the regulation instruction. If it is greater than the abnormal point threshold, update the regulation direction to the opposite direction of the preliminary regulation direction, combine the current abnormal device operation parameter value, and update according to the operation standard range and regulated step size length of the abnormal device operation parameters to obtain an updated regulation instruction sequence, and perform abnormal monitoring on the updated regulation instruction sequence; otherwise continue to execute the next regulation instruction in the regulation instruction sequence until the regulation instruction sequence of the current abnormal device operation parameter is executed; If the deviation degree of the abnormal production parameter is less than or equal to the stop threshold, stop the regulation of the device operation parameters.

9. The device parameter collaborative control system according to claim 8, wherein The specific steps for performing abnormal monitoring on the updated regulation instruction sequence include: After executing the regulation instruction in the updated regulation instruction sequence, obtain the deviation degree of the abnormal production parameter. If the deviation degree is alleviated, continue to execute the next regulation instruction in the updated regulation instruction sequence until the updated regulation instruction sequence of the current abnormal production parameter is executed completely; If the deviation degree is not alleviated, mark the abnormal point of the updated regulation instruction, conduct monitoring of the updated regulation instruction, count the number of abnormal points of the updated regulation instruction. If it is greater than the abnormal point threshold, stop the regulation and give an early warning of regulation abnormality. Otherwise, continue to execute the next regulation instruction in the updated regulation instruction sequence until the updated regulation instruction sequence of the current abnormal production parameter is executed completely; When the regulation instruction sequences of all the operating parameters of the abnormal execution devices are executed, if the deviation degree of the abnormal production parameter is greater than the stop threshold, give an early warning of regulation abnormality.

10. A device parameter collaborative control method, which is implemented based on the device parameter collaborative control system described in any one of claims 1-9, characterized in that It includes the following steps: Step S1: Determine the initial production parameter value based on the production task of the target device, and collect the real-time production parameter value during the production task execution of the target device. Through multi-level comparative analysis of the real-time production parameter and its deviation degree, conduct sudden disturbance monitoring to judge whether there is a sudden disturbance in the target device; Step S2: When it is detected that there is a sudden disturbance in the target device, construct a parameter abnormal propagation path map to locate the abnormal production parameter through Bayes' theorem, and based on the abnormal production parameter, inversely infer and locate the abnormal device operating parameter that causes the sudden disturbance; Step S3: Generate the regulation instructions for the abnormal device operating parameter in sequence according to the arrangement order of the abnormal device operating parameter that causes the sudden disturbance, and dynamically update the regulation instructions for the abnormal device operating parameter based on the real-time deviation degree of the abnormal production parameter value; Step S4: After the target device completes the production task, collect the quality parameter value, conduct quality inspection on the product. If the product quality inspection is unqualified, give an early warning of product quality and generate a product quality production report.

Images