Valve optimization control method and system based on machine learning

Abstract

The invention belongs to the technical field of valve control, and provides a valve optimization control method and system based on machine learning, and the method comprises the steps: carrying out the abnormality analysis of the opening and closing response time when fluid exists, and judging whether the opening and closing response time of a valve is abnormal or not; if the valve opening and closing response time is abnormal, combining the instantaneous flow velocity corresponding to the valve opening and closing response time, analyzing the possibility that the instantaneous flow velocity affects the valve opening and closing response time, and if the possibility that the instantaneous flow velocity affects the valve opening and closing response time is large, judging whether the influence of the instantaneous flow velocity on the valve opening and closing response time is regular or not; constructing a prediction model, and judging the prediction accuracy of the prediction model; according to the prediction accuracy judgment result of the prediction model, an opening and closing response time adjusting value is determined, the opening and closing response time of the valve is adjusted, flow speed control is quantitatively guided through the opening and closing response time adjusting value, and model self-adaptive evolution is driven through dynamic feedback of monitoring data.

Description

Technical Field

[0001] The present invention belongs to the technical field of valve control, and particularly relates to a valve optimization control method and system based on machine learning. Background Art

[0002] In industrial process control, as a key actuator component of the fluid system, the stability of the opening and closing response time of the valve directly affects the safety and efficiency of the system.

[0003] However, there are obvious deficiencies in the existing technology in terms of valve response time control: the ability of multi-dimensional data correlation analysis is insufficient, and it is difficult to combine the instantaneous flow rate to locate the hydrodynamic causes of abnormal response time; the prediction model lacks a dynamic adaptation mechanism, and the accuracy drops significantly in the face of equipment aging or working condition changes; the adjustment strategy relies on manual experience, and fine control has not been achieved according to the law between the opening and closing response time of the valve and the instantaneous flow rate.

[0004] Therefore, the present invention provides a valve optimization control method and system based on machine learning. Summary of the Invention

[0005] In order to make up for the deficiencies of the existing technology and solve at least one of the technical problems proposed in the background art.

[0006] In a first aspect, the present invention provides a valve optimization control method based on machine learning, including: Construct an opening and closing response time analysis library according to the opening and closing response time of the valve within the historical period, and judge whether the opening and closing response time of the valve is abnormal by analyzing the abnormality of the opening and closing response time when there is fluid; If the opening and closing response time of the valve is abnormal, combine the instantaneous flow rate corresponding to the opening and closing response time of the valve, analyze the possibility that the instantaneous flow rate affects the opening and closing response time of the valve, and if the possibility that the instantaneous flow rate affects the opening and closing response time is large, judge whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular; Construct a prediction model according to the analysis result of whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular, output the adjusted instantaneous flow rate range, and perform analysis to judge the prediction accuracy of the prediction model; Determine the opening and closing response time adjustment value according to the prediction accuracy judgment result of the prediction model, and adjust the opening and closing response time of the valve.

[0007] Preferably, the specific process of judging whether the opening and closing response time of the valve is abnormal is: Divide the historical period into several historical time points at equal time intervals, export the opening and closing response times of the valve at all historical time points from the opening and closing response time analysis library, analyze the opening and closing response times at all historical time points within the historical period, determine the abnormal response ratio, and if it is greater than or equal to the abnormal response ratio threshold, it indicates that the opening and closing response time of the valve is abnormal.

[0008] Preferably, the process of determining the abnormal response ratio is as follows: Subtract the opening and closing response time of the valve at the historical time point from the standard value of the opening and closing response time, and take the absolute value of the difference to obtain the response time deviation value. If it is greater than or equal to the response time deviation standard value, mark the corresponding historical time point as an abnormal response time point; Calculate the proportion of the number of abnormal response time points among all historical time points, which is denoted as the abnormal response ratio.

[0009] Preferably, the specific process of analyzing the possibility that the instantaneous flow rate affects the opening and closing response time of the valve is as follows: Extract the instantaneous flow rate corresponding to the opening and closing response time of the abnormal response time point, and use the Spearman coefficient formula to take the absolute value to calculate the monotonic correlation value between the opening and closing response time of the valve and the instantaneous flow rate. If it is greater than or equal to the monotonic correlation value threshold, it indicates that the instantaneous flow rate has a high possibility of affecting the opening and closing response time.

[0010] Preferably, the specific process of determining whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular is as follows: Use the least squares method to perform fitting analysis on the opening and closing response time of the abnormal response time point and the corresponding instantaneous flow rate, calculate the slope of the fitting line and the intercept of the fitting line, and construct a fitting model of the opening and closing response time and the instantaneous flow rate; Use the goodness-of-fit formula to calculate the regularity characterization value of the fitting line between the opening and closing response time and the instantaneous flow rate. If it is greater than or equal to the regularity characterization threshold, it indicates that the influence of the instantaneous flow rate on the opening and closing response time is linearly regular.

[0011] Preferably, the process of determining the adjusted instantaneous flow rate interval is as follows: Input the standard value of the opening and closing response time of the valve into the prediction model to obtain the instantaneous flow rate when the opening and closing response time of the valve reaches the standard value of the opening and closing response time, which is denoted as the arrival instantaneous flow rate, and perform a difference process with the current instantaneous flow rate to obtain the adjusted instantaneous flow rate. Denote the interval between the adjusted instantaneous flow rates as the adjusted instantaneous flow rate interval.

[0012] Preferably, the process of judging the prediction accuracy of the prediction model is as follows: Within the adjustable instantaneous flow rate range, a preset monitoring range is set, and within the monitoring range, the opening and closing response time of the valve and the instantaneous flow rate are monitored in real time. By comparing and analyzing the different instantaneous flow rates and predicted instantaneous flow rates monitored in real time within the monitoring range, an asynchronous ratio value is determined. If it is greater than or equal to the asynchronous ratio threshold, it indicates that the prediction model is accurate; otherwise, it indicates that the prediction model is inaccurate.

[0013] Preferably, the process for determining the asynchronous ratio value is as follows: Based on any instantaneous flow rate within the preset monitoring range; if the difference between the instantaneous flow rate within the preset monitoring range and the predicted instantaneous flow rate is calculated and then the absolute value is taken, the relative deviation of the instantaneous flow rate is obtained. Compare the relative deviation of the instantaneous flow rate with the standard value of the relative deviation of the instantaneous flow rate. If it is greater than or equal to the standard value of the relative deviation of the instantaneous flow rate, the corresponding predicted instantaneous flow rate is recorded as the asynchronous instantaneous flow rate; calculate the proportion of the number of asynchronous instantaneous flow rates within the monitoring range to obtain the asynchronous ratio value.

[0014] Preferably, the process for determining the adjustment value of the opening and closing response time is as follows: If the prediction model is inaccurate, the instantaneous flow rate and the corresponding opening and closing response time monitored within the monitoring range are supplemented to the training dataset of the prediction model, the prediction model is corrected, and the adjustable instantaneous flow rate range is re-output and used as the analysis range. If the prediction model is accurate, for the already trained prediction model, the adjustable instantaneous flow rate range output is used as the analysis range. Extract the predicted instantaneous flow rate and the corresponding opening and closing response time within the analysis range and perform fitting analysis again. If it conforms to a linear pattern, the slope of the fitting line is used as the adjustment coefficient. If it does not conform to a linear pattern, the analysis range is divided into several analysis segments with equal time intervals. The difference between the maximum opening and closing response time and the minimum opening and closing response time within the analysis segment is calculated to obtain the segment time difference. The difference between the instantaneous flow rate corresponding to the maximum opening and closing response time and the flow rate corresponding to the minimum opening and closing response time within the analysis segment is calculated to obtain the instantaneous flow rate difference. The instantaneous flow rate difference is divided by the segment time difference to obtain the change rate. The change rates of all analysis segments are integrated into a change rate sequence, and the maximum change rate in the change rate sequence is extracted as the adjustment coefficient. Take the absolute value of the difference between the current instantaneous flow rate and the predicted instantaneous flow rate to obtain the change value of the instantaneous flow rate, and multiply the change value of the instantaneous flow rate by the adjustment coefficient to obtain the adjustment value of the opening and closing response time.

[0015] In a second aspect, the present invention also provides a valve optimization control system based on machine learning. The system includes: Abnormality judgment module: Based on the opening and closing response times of the valve within a historical period, construct an analysis library for opening and closing response times, and determine whether the opening and closing response time of the valve is abnormal by performing abnormality analysis on the opening and closing response times when there is fluid; Regularity analysis module: If the opening and closing response time of the valve is abnormal, combine the instantaneous flow rate corresponding to the opening and closing response time of the valve to analyze the possibility that the instantaneous flow rate affects the opening and closing response time of the valve. If the possibility that the instantaneous flow rate affects the opening and closing response time is large, determine whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular; Model optimization module: According to the analysis result of whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular, construct a prediction model, output the adjusted instantaneous flow rate range, and perform analysis to judge the prediction accuracy of the prediction model; Regulation analysis module: According to the judgment result of the prediction accuracy of the prediction model, determine the opening and closing response time adjustment value and adjust the opening and closing response time of the valve.

[0016] The beneficial effects of the present invention are as follows: 1. By constructing an analysis library for opening and closing response times driven by historical data, the present invention ensures data reliability, uses timestamps to map instantaneous flow rates and response times, and realizes multi-dimensional data correlation analysis; sets standard values based on historical experience and quantifies the degree of abnormality through the abnormal response ratio, which can effectively identify the overall abnormal state of the valve response time; introduces the Spearman coefficient to analyze the monotonic correlation, and can accurately judge the possibility of the flow rate affecting the response time; through the least squares method to fit a linear model and combined with the goodness-of-fit test, the type of law of the flow rate influence can be further clarified, and the slope of the fitted line provides a quantitative basis for adjusting the opening and closing time of the valve.

[0017] 2. Based on the law of flow rate influence, the present invention intelligently selects a linear regression or machine learning model, combines the division of the training set - validation set and the optimization of the mean square error, and significantly improves the accuracy of response time prediction; by setting a monitoring interval with an interval length smaller than the adjustment interval, the quantification evaluation of the proximity between the predicted value and the real-time flow rate is realized, and the asynchronous ratio threshold comparison mechanism effectively identifies the model deviation, providing a clear trigger condition for subsequent model iteration; further innovatively proposes a dynamic adjustment coefficient calculation strategy, which not only supports the direct application of the slope in the linear scenario, but also extracts the section change rate in the non-linear scenario, realizing refined control under complex working conditions; not only quantitatively guides the flow rate control through the opening and closing response time adjustment value, but also drives the model to adaptively evolve through the dynamic feedback of the monitoring data, significantly improving the intelligence, response speed and long-term stability of the valve control system. Description of the Drawings

[0018] The present invention will be further described below with reference to the accompanying drawings.

[0019] Figure 1 is a flowchart of the steps of a valve optimization control method based on machine learning according to an embodiment of the present invention; Figure 2 is a system block diagram of a valve optimization control system based on machine learning according to an embodiment of the present invention. Detailed implementation manners

[0020] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific implementation manners.

[0021] Embodiment 1 Please refer to Figure 1 As shown, a valve optimization control method based on machine learning according to an embodiment of the present invention includes the following steps: Step 1: According to the opening and closing response time of the valve within a historical period, construct an opening and closing response time analysis library, and judge whether the opening and closing response time of the valve is abnormal by performing anomaly analysis on the opening and closing response time when there is fluid; It should be noted that the historical period includes but is not limited to three months, half a year, and one year. The opening and closing response time refers to the complete action time of the valve from fully open to fully closed or from fully closed to fully open; Obtain the opening and closing response time of the valve within the historical period from the data acquisition and monitoring control system and the distributed control system, collect the instantaneous flow rate data when the valve opens and closes through the flow rate sensor, and perform data preprocessing on the collected data, including removing outliers using the IQR method, interpolating and complementing missing data, using Min-Max normalization processing, mapping the opening and closing response time of the valve and the instantaneous flow rate through the time stamp, and storing the opening and closing response time and instantaneous flow rate data in the MySQL database to construct an opening and closing response time analysis library; It can be understood that within the historical period, the fluid type and valve type are fixed, and data is collected at a fixed sampling frequency; Divide the historical period into several historical time points at equal time intervals, export the opening and closing response time of the valve at all historical time points from the opening and closing response time analysis library, perform a difference operation on the opening and closing response time of the valve at the historical time point and the opening and closing response time standard value, and take the absolute value of the difference to obtain the response time deviation value; Compare the response time deviation value with the response time deviation standard value: If the response time deviation value is greater than or equal to the response time deviation standard value, mark the corresponding historical time point as a response abnormal time point; If the response time deviation value is less than the response time deviation standard value, mark the corresponding historical time point as a response normal time point; It is understood that the abnormal response time point indicates that the valve opening and closing response time deviates from the standard value of the opening and closing response time, reflecting the abnormality of the valve opening and closing response time. The normal response time point indicates that the valve opening and closing response time does not deviate from the standard value of the opening and closing response time, reflecting the normal opening and closing response time of the valve. The standard value of the opening and closing response time and the standard value of the response time deviation are both set by those skilled in the art based on historical experience. Count the number of abnormal response time points in all historical time points, and calculate the ratio of the number of abnormal response time points in all historical time points, which is recorded as the abnormal response ratio; In some embodiments, the abnormal response ratio is compared to an abnormal response ratio threshold: If the abnormal response ratio is greater than or equal to the abnormal response ratio threshold, it means that the opening and closing response time of the valve is abnormal; If the abnormal response ratio is less than the abnormal response ratio threshold, it means that the opening and closing response time of the valve is normal; Step 2: If the valve opening and closing response time is abnormal, analyze the possibility that the instantaneous flow velocity affects the valve opening and closing response time based on the instantaneous flow velocity corresponding to the valve opening and closing response time. If the possibility that the instantaneous flow velocity affects the opening and closing response time is high, determine whether the effect of the instantaneous flow velocity on the valve opening and closing response time is regular. Extract the instantaneous flow velocity corresponding to the opening and closing response time at the abnormal response time point using the Spearman coefficient formula: Take the absolute value to calculate the monotonic correlation between the valve opening and closing response time and the instantaneous flow rate, where, represents the monotonic correlation value between the valve opening and closing response time and the instantaneous flow rate, n represents the number of samples, It represents the difference between the rank of the instantaneous flow velocity and the rank of the opening and closing response time in the i-th sample; If the monotonic correlation value between the valve opening and closing response time and the instantaneous flow velocity is greater than or equal to the monotonic correlation value threshold, it means that the instantaneous flow velocity is likely to have an impact on the opening and closing response time; If the monotonic correlation value between the valve opening and closing response time and the instantaneous flow velocity is less than the monotonic correlation value threshold, it means that the possibility of the instantaneous flow velocity affecting the opening and closing response time is small; Based on the possibility that the instantaneous flow velocity has a great impact on the opening and closing response time, the least squares method is used to fit the opening and closing response time and the corresponding instantaneous flow velocity at the time point of the abnormal response. The fitting line equation is y=kx+b, where k represents the slope of the fitting line, b represents the intercept of the fitting line, x represents the instantaneous flow velocity, and y represents the opening and closing response time. By formula: Calculate the slope k of the fitted line, where represents the i-th instantaneous flow velocity sample value, represents the average value of the instantaneous flow velocity samples, represents the sample value of the i-th opening and closing response time, represents the average value of the opening and closing response time samples; through the formula: calculate the intercept of the fitting straight line; It should be noted that since the time stamp maps the valve opening and closing response time to the instantaneous flow velocity, so and have the same meaning of i; According to the slope of the fitting straight line and the intercept of the fitting straight line calculated, construct a fitting model of the opening and closing response time and the instantaneous flow velocity; Use the goodness-of-fit formula calculate the law characterization value of the fitting straight line between the opening and closing response time and the instantaneous flow velocity , where SSE represents the sum of squared residuals and SST represents the total sum of squares; It can be understood that the sum of squared residuals characterizes the error between the predicted value and the actual value of the fitting model, and the total sum of squares characterizes the total variation degree of the data itself; Compare the law characterization value with the law characterization threshold: If the law characterization value is greater than or equal to the law characterization threshold, it means that the influence of the instantaneous flow velocity on the opening and closing response time presents a linear law; If the law characterization value is less than the law characterization threshold, it means that the influence of the instantaneous flow velocity on the opening and closing response time presents a non-linear law; It should be noted that the necessity of using the least squares method to perform fitting analysis on the opening and closing response time and the instantaneous flow velocity of the response abnormal time points is that through the slope of the fitting straight line, the unit influence of the valve instantaneous flow velocity on the opening and closing response time can be clarified, providing a reliable support for the subsequent adjustment of the valve opening and closing time; The technical solution of this embodiment is: The present invention constructs an opening and closing response time analysis library driven by historical data to ensure data reliability, uses the time stamp to map the instantaneous flow velocity and the response time to realize multi-dimensional data correlation analysis; sets the standard value based on historical experience and quantifies the abnormal degree through the abnormal response ratio, and can effectively identify the overall abnormal state of the valve response time; introduces the Spearman coefficient to analyze the monotonic correlation, and can accurately judge the possibility of the influence of the flow velocity on the response time; through the least squares method to fit the linear model and combine the goodness-of-fit test, the law type of the influence of the flow velocity can be further clarified, and the slope of the fitting straight line provides a quantitative basis for the adjustment of the valve opening and closing time.

[0022] Embodiment 2 Please refer to Figure 1 as shown, a valve optimization control method based on machine learning described in an embodiment of the present invention includes the following steps: Step 3: Based on the analysis result of whether the influence of the instantaneous flow rate on the valve opening and closing response time is regular, construct a prediction model, output the adjusted instantaneous flow rate range, and conduct an analysis to judge the prediction accuracy of the prediction model; Integrate the opening and closing response times and the corresponding instantaneous flow rates at all abnormal response time points into a data set, and divide it into a training set (70%) and a validation set (30%) in chronological order. If the influence of the instantaneous flow rate on the opening and closing response time is linear, construct a prediction model based on the linear regression model. If the influence of the instantaneous flow rate on the opening and closing response time is non-linear, construct a prediction model based on the machine learning model. Use the training data set to train the selected prediction model to obtain the trained prediction model. Use the validation set to verify the trained prediction model, use the mean squared error as the loss function, and use Adam as the optimizer; Input the standard value of the valve opening and closing response time into the prediction model to obtain the instantaneous flow rate when the valve opening and closing response time reaches the standard value of the opening and closing response time, denoted as the arrival instantaneous flow rate. Take the difference between the arrival instantaneous flow rate and the current instantaneous flow rate to obtain the adjusted instantaneous flow rate, and denote the interval between the adjusted instantaneous flow rates as the adjusted instantaneous flow rate range; Within the adjusted instantaneous flow rate range, preset a monitoring range. In the monitoring range, monitor the opening and closing response time and the instantaneous flow rate of the valve in real time. By comparing and analyzing the different instantaneous flow rates monitored in real time within the monitoring range with the predicted instantaneous flow rate, judge the degree of approximation between the instantaneous flow rate monitored in real time within the adjusted instantaneous flow rate range and the predicted instantaneous flow rate; Among them, the interval length corresponding to the preset monitoring range is less than the interval length corresponding to the adjusted instantaneous flow rate range; It should be noted that the purpose of presetting the monitoring range within the adjusted instantaneous flow rate range is to analyze the prediction accuracy by monitoring in real time whether the instantaneous flow rate within the preset monitoring range is close to the predicted instantaneous flow rate; Based on any instantaneous flow rate within the preset monitoring range; if the difference between the instantaneous flow rate within the preset monitoring range and the predicted instantaneous flow rate is processed and then the absolute value is taken, the relative deviation of the instantaneous flow rate is obtained; Compare the relative deviation of the instantaneous flow rate with the standard value of the relative deviation of the instantaneous flow rate. If the relative deviation of the instantaneous flow rate is greater than or equal to the standard value of the relative deviation of the instantaneous flow rate, record the corresponding predicted instantaneous flow rate as the asynchronous instantaneous flow rate; if the relative deviation of the instantaneous flow rate is less than the standard value of the relative deviation of the instantaneous flow rate, record the corresponding predicted instantaneous flow rate as the synchronous instantaneous flow rate; It should be noted that the standard value of the relative deviation of the instantaneous flow rate is set by those skilled in the art according to historical experience; Count the number of asynchronous instantaneous flow rates within the monitoring range, calculate the proportion of the number of asynchronous instantaneous flow rates within the monitoring range, and obtain the asynchronous proportion value; In some embodiments, the asynchronous ratio value is compared with the asynchronous ratio threshold; If the asynchronous ratio value is greater than or equal to the asynchronous ratio threshold, it indicates that the instantaneous flow rate change trend of the valve within the monitoring interval is similar to the predicted instantaneous flow rate change trend, indicating that the prediction model is accurately predicted; If the asynchronous ratio value is less than the asynchronous ratio threshold, it indicates that the instantaneous flow rate change trend of the valve within the monitoring interval is not similar to the predicted instantaneous flow rate change trend, indicating that the prediction model is not accurately predicted; Step 4: Determine the opening and closing response time adjustment value according to the prediction accuracy judgment result of the prediction model, and adjust the opening and closing response time of the valve; If the prediction model is not accurately predicted, the instantaneous flow rate and the corresponding opening and closing response time monitored in the monitoring interval are supplemented to the training data set of the prediction model, the prediction model is corrected, and the adjusted instantaneous flow rate interval is re-output and used as the analysis interval; If the prediction model is accurately predicted, for the already trained prediction model, the adjusted instantaneous flow rate interval output is used as the analysis interval; Extract the predicted instantaneous flow rate and the corresponding opening and closing response time within the analysis interval, and perform linear fitting analysis again; If it conforms to the linear law, the slope of the fitting line obtained by the fitting analysis again is used as the adjustment coefficient; If it does not conform to the linear law, the analysis interval is divided into several analysis segments with equal time intervals. The difference between the maximum opening and closing response time and the minimum opening and closing response time within the analysis segment is calculated to obtain the segment time difference. The difference between the instantaneous flow rate corresponding to the maximum opening and closing response time and the flow rate corresponding to the minimum opening and closing response time within the analysis segment is calculated to obtain the instantaneous flow rate difference. The instantaneous flow rate difference is divided by the segment time difference to obtain the change rate. The change rates of all analysis segments are integrated into a change rate sequence, and the maximum change rate in the change rate sequence is extracted as the adjustment coefficient; According to the adjustment coefficient, the absolute value of the difference between the current instantaneous flow rate and the predicted instantaneous flow rate is taken to obtain the instantaneous flow rate change value, and the instantaneous flow rate change value is multiplied by the adjustment coefficient to obtain the opening and closing response time adjustment value; Adjust the opening and closing response time of the valve according to the opening and closing response time adjustment value. Accurate response time can avoid the fluid water hammer effect when the valve opens and closes; The technical solution of this embodiment is as follows: based on a model with a suitable flow rate influence law, the present invention combines training set-validation set division with mean square error optimization to significantly improve the accuracy of response time prediction; by setting a monitoring interval with an interval length shorter than the adjustment interval, a quantitative evaluation of the approximation between the predicted value and the real-time flow rate is achieved, and the asynchronous proportional threshold comparison mechanism effectively identifies model deviations, providing clear trigger conditions for subsequent model iterations; further innovatively proposes a dynamic adjustment coefficient calculation strategy, which not only supports direct application of the slope in linear scenarios, but also realizes refined regulation under complex working conditions through extraction of segment change rates in nonlinear scenarios; not only does it quantitatively guide flow rate control through the opening and closing response time adjustment value, but it also drives the adaptive evolution of the model through dynamic feedback of monitoring data, significantly improving the intelligence, response speed and long-term stability of the valve control system.

[0023] Example 3 See also Figure 2 As shown, a valve optimization control system based on machine learning according to an embodiment of the present invention includes the following modules: Abnormality judgment module: Based on the valve opening and closing response time in the historical period, an opening and closing response time analysis library is built. By performing abnormal analysis on the opening and closing response time when fluid is present, it is determined whether the valve opening and closing response time is abnormal; Regularity Analysis Module: If the valve opening and closing response time is abnormal, the module analyzes the possibility that the instantaneous flow velocity affects the valve opening and closing response time in combination with the instantaneous flow velocity corresponding to the valve opening and closing response time. If the possibility of the instantaneous flow velocity affecting the valve opening and closing response time is high, the module determines whether the influence of the instantaneous flow velocity on the valve opening and closing response time is regular. Model optimization module: Based on the analysis results of whether the impact of instantaneous flow velocity on valve opening and closing response time is regular, a prediction model is constructed, the output adjustment instantaneous flow velocity range is output, and analysis is performed to determine the prediction accuracy of the prediction model; Control and analysis module: Based on the prediction accuracy of the prediction model, the opening and closing response time adjustment value is determined and the opening and closing response time of the valve is adjusted.

[0024] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A valve optimization control method based on machine learning, characterized in that: Including: Construct an opening and closing response time analysis library based on the opening and closing response times of the valve within a historical period. By performing anomaly analysis on the opening and closing response times when there is fluid, determine whether the opening and closing response time of the valve is abnormal; If the opening and closing response time of the valve is abnormal, combine the instantaneous flow rate corresponding to the opening and closing response time of the valve to analyze the possibility that the instantaneous flow rate affects the opening and closing response time of the valve. If the possibility that the instantaneous flow rate affects the opening and closing response time is large, determine whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular; Based on the analysis result of whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular, construct a prediction model, output the adjusted instantaneous flow rate range, and perform analysis to judge the prediction accuracy of the prediction model; According to the judgment result of the prediction accuracy of the prediction model, determine the opening and closing response time adjustment value and adjust the opening and closing response time of the valve.

2. The valve optimization control method based on machine learning according to claim 1, wherein: The specific process of determining whether the opening and closing response time of the valve is abnormal is as follows: Divide the historical period into several historical time points at equal time intervals. Export the opening and closing response times of the valve at all historical time points from the opening and closing response time analysis library, analyze the opening and closing response times at all historical time points within the historical period, determine the abnormal response ratio. If it is greater than or equal to the abnormal response ratio threshold, it indicates that the opening and closing response time of the valve is abnormal.

3. The valve optimization control method based on machine learning according to claim 2, characterized in that: The process of determining the abnormal response ratio is as follows: Subtract the opening and closing response time of the valve at the historical time point from the standard value of the opening and closing response time, and take the absolute value of the difference to obtain the response time deviation value. If it is greater than or equal to the response time deviation standard value, record the corresponding historical time point as a response abnormal time point; Calculate the proportion of the number of response abnormal time points among all historical time points, which is recorded as the abnormal response ratio.

4. The valve optimization control method based on machine learning according to claim 2, wherein: The specific process of analyzing the possibility that the instantaneous flow rate affects the opening and closing response time of the valve is as follows: Extract the instantaneous flow rate corresponding to the opening and closing response time of the response abnormal time point, and use the Spearman coefficient formula to take the absolute value to calculate the monotonic correlation value between the opening and closing response time of the valve and the instantaneous flow rate. If it is greater than or equal to the monotonic correlation value threshold, it indicates that the possibility that the instantaneous flow rate affects the opening and closing response time is large.

5. The valve optimization control method based on machine learning according to claim 2, characterized in that: The specific process of determining whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular is as follows: Use the least squares method to perform fitting analysis on the opening and closing response time of the response abnormal time point and the corresponding instantaneous flow rate, calculate the slope of the fitting line and the intercept of the fitting line, and construct a fitting model of the opening and closing response time and the instantaneous flow rate; Use the goodness-of-fit formula to calculate the regularity characterization value of the fitting line between the opening and closing response time and the instantaneous flow rate. If it is greater than or equal to the regularity characterization threshold, it indicates that the influence of the instantaneous flow rate on the opening and closing response time is linearly regular.

6. The valve optimization control method based on machine learning according to claim 3, wherein: The process of determining the adjusted instantaneous flow rate range is as follows: Input the standard value of the opening and closing response time of the valve into the prediction model to obtain the instantaneous flow rate when the opening and closing response time of the valve reaches the standard value of the opening and closing response time, which is recorded as the arrival instantaneous flow rate, and perform a difference process with the current instantaneous flow rate to obtain the adjusted instantaneous flow rate. Record the interval between the adjusted instantaneous flow rates as the adjusted instantaneous flow rate range.

7. The valve optimization control method based on machine learning according to claim 6, characterized in that: The process of judging the prediction accuracy of the prediction model is as follows: Within the adjusted instantaneous flow rate range, a preset monitoring range is set. The opening and closing response time of the valve and the instantaneous flow rate are monitored in real time within the monitoring range. By comparing and analyzing the different instantaneous flow rates and the predicted instantaneous flow rates monitored in real time within the monitoring range, an asynchronous ratio value is determined. If it is greater than or equal to the asynchronous ratio threshold, it indicates that the prediction model is accurately predicted; otherwise, it indicates that the prediction model is inaccurately predicted.

8. A valve optimization control method based on machine learning according to claim 7, characterized in that: The process of determining the asynchronous ratio value is as follows: Based on any instantaneous flow rate within the preset monitoring range; if the difference between the instantaneous flow rate within the preset monitoring range and the predicted instantaneous flow rate is calculated and then the absolute value is taken, the relative deviation of the instantaneous flow rate is obtained. The relative deviation of the instantaneous flow rate is compared with the standard value of the relative deviation of the instantaneous flow rate. If it is greater than or equal to the standard value of the relative deviation of the instantaneous flow rate, the corresponding predicted instantaneous flow rate is recorded as the asynchronous instantaneous flow rate; the proportion of the number of asynchronous instantaneous flow rates within the monitoring range is calculated to obtain the asynchronous ratio value.

9. The valve optimization control method based on machine learning according to claim 7, characterized in that: The process of determining the adjustment value of the opening and closing response time is as follows: If the prediction model is inaccurately predicted, the instantaneous flow rate and the corresponding opening and closing response time monitored within the monitoring range are supplemented to the training dataset of the prediction model, the prediction model is corrected, and the adjusted instantaneous flow rate range is output again and used as the analysis range. If the prediction model is accurately predicted, for the already trained prediction model, the output adjusted instantaneous flow rate range is used as the analysis range. The predicted instantaneous flow rate and the corresponding opening and closing response time within the analysis range are extracted and subjected to fitting analysis again. If it conforms to the linear law, the slope of the fitting line is used as the adjustment coefficient. If it does not conform to the linear law, the analysis range is divided into several analysis sections with equal time intervals. The difference between the maximum opening and closing response time and the minimum opening and closing response time within the analysis section is calculated to obtain the section time difference. The difference between the instantaneous flow rate corresponding to the maximum opening and closing response time and the flow rate corresponding to the minimum opening and closing response time within the analysis section is calculated to obtain the instantaneous flow rate difference. The instantaneous flow rate difference is divided by the section time difference to obtain the change rate. The change rates of all analysis sections are integrated into a change rate sequence, and the maximum change rate in the change rate sequence is extracted as the adjustment coefficient. The absolute value of the difference between the current instantaneous flow rate and the predicted instantaneous flow rate is taken to obtain the instantaneous flow rate change value. The instantaneous flow rate change value is multiplied by the adjustment coefficient to obtain the adjustment value of the opening and closing response time.

10. A valve optimization control system based on machine learning, characterized in that, The system is used to execute the method described in any one of the above claims 1-9. The system includes: Anomaly judgment module: Based on the opening and closing response time of the valve within the historical period, an opening and closing response time analysis library is constructed. By performing anomaly analysis on the opening and closing response time when there is fluid, it is judged whether the opening and closing response time of the valve is abnormal. Regularity analysis module: If the opening and closing response time of the valve is abnormal, in combination with the instantaneous flow rate corresponding to the opening and closing response time of the valve, the possibility of the instantaneous flow rate affecting the opening and closing response time of the valve is analyzed. If the possibility of the instantaneous flow rate affecting the opening and closing response time is large, it is judged whether the influence of the instantaneous flow rate on the opening and closing response time of the valve is regular. Model optimization module: Based on the analysis result of whether the influence of the instantaneous flow rate on the valve opening and closing response time is regular, construct a prediction model, output the adjusted instantaneous flow rate range, and conduct an analysis to judge the prediction accuracy of the prediction model; Regulation analysis module: Based on the judgment result of the prediction accuracy of the prediction model, determine the adjustment value of the opening and closing response time, and adjust the opening and closing response time of the valve.

Images