Method for monitoring water content in the dehydration process of oleic acid for the preparation of diesel antiwear agent

Abstract

The present invention relates to the technical field of data processing, and more specifically, to a method for monitoring the water content of an oleic acid dehydration process for preparing a diesel antiwear agent, comprising: real-time acquisition of a data sequence of water content data of an oleic acid dehydration process; curve fitting of each data point in the data sequence and each data point within a preset local range of each data point; determining the degree of abnormal performance of each data point in the data sequence; correcting the degree of abnormal performance by the difference between a reference value and an actual value of each data point in the data sequence; constructing a scatter plot to determine the accuracy of an RRCF model; updating the RRCF model according to the accuracy; and performing detection using the updated RRCF model to complete the water content monitoring of the oleic acid dehydration process. The present invention can more accurately determine when the model needs to be updated, avoid unnecessary updates that lead to waste of computing resources, and simultaneously improve the detection efficiency of water content data in the oleic acid dehydration process.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a method for monitoring the water content in the dehydration process of oleic acid used in the preparation of diesel antiwear agent. Background Art

[0002] Diesel anti-wear agent, also known as diesel lubricant improver. In the preparation process of diesel anti-wear agent, especially in reactions involving fatty acids and other chemicals, the presence of water may have a negative impact on the chemical reaction. Because oleic acid dehydration can remove water from the reaction system, thereby avoiding the interference of water on the reaction, oleic acid dehydration is an important step to improve the efficiency of the reaction and the quality of the product. Therefore, in order to ensure that the prepared anti-wear agent has good performance, it is necessary to conduct real-time abnormal detection of the water content in the oleic acid dehydration process to promptly discover and solve problems.

[0003] The patent document with announcement number CN116304524B discloses a soil moisture monitoring method, equipment, storage medium and device, the method comprising: determining a multivariate linear regression model for soil moisture monitoring based on the multivariate linear relationship between a preset soil moisture monitoring model and an influencing factor of soil moisture; performing Gaussian process regression analysis based on the observed values ​​of the preset soil moisture monitoring model and the monitored values ​​of the multivariate linear regression model to obtain a Gaussian process regression model for soil moisture monitoring; optimizing the Gaussian process regression model based on Bayesian theorem to obtain an optimal soil moisture monitoring model and monitoring soil moisture according to the optimal soil moisture monitoring model.

[0004] However, the above document is for monitoring soil moisture content, and cannot be used for monitoring moisture content during oleic acid dehydration. The Robust Random Cut Forest (RRCF) algorithm can be used for detection to complete the monitoring of moisture content during oleic acid dehydration. The RRCF algorithm is an anomaly detection algorithm for dynamic data volume. The algorithm dynamically maintains the distribution of random segmentation trees on the data stream, and judges its degree of abnormality based on the impact of the insertion or deletion of new sample points on the complexity of the segmentation tree. When using the RRCF model to detect anomalies in the moisture content during oleic acid dehydration, since the moisture content data is time series data, the data distribution of moisture content will inevitably change with the passage of time. Therefore, it is necessary to regularly update the RRCF model to adapt to the new data distribution. The traditional regular update strategy may lead to unnecessary updates when the data distribution is stable, resulting in a waste of computing resources, and may not be able to respond to the rapid changes in data distribution in a timely manner, resulting in inaccurate detection of moisture anomalies, affecting the preparation efficiency of diesel anti-wear agent. Summary of the invention

[0005] In order to solve the problem that the fixed update of the RRCF model causes unnecessary updates when the data distribution is stable, resulting in a waste of computing resources, and the inability to respond to the rapid changes in the data distribution in a timely manner, resulting in inaccurate water content detection and affecting the preparation efficiency of the diesel antiwear agent, the present invention proposes a method for monitoring the water content in the oleic acid dehydration process for the preparation of the diesel antiwear agent, the method comprising:

[0006] The water content data of the oleic acid dehydration process collected in real time are cleaned to obtain a water content data sequence; curve fitting is performed on each data point in the data sequence and each data point within a preset local range of each data point, and the slope value of each data point is planned according to the slope value of each data point on the fitting curve, and the slope value after planning is positively correlated with the slope value of the data point; the abnormal performance degree of each data point in the data sequence is determined, and the abnormal performance degree is positively correlated with the difference between the mean of the slope values ​​of each data point in the data sequence after planning and the mean of the slope values ​​of each data point in the preset local range of the data point; through the reference value and the actual value of each data point in the data sequence The abnormal performance degree is corrected according to the difference, and the corrected abnormal performance degree is positively correlated with the abnormal performance degree and negatively correlated with the difference; the abnormal score of each data point in the data sequence is obtained through the RRCF model, and a scatter plot is constructed with the abnormal performance degree of each data point as the vertical coordinate and the abnormal score as the horizontal coordinate to determine the accuracy of the RRCF model, and the accuracy is negatively correlated with the mean value of the distance from each data point in the scatter plot to the regression curve of the scatter plot; according to the accuracy, the RRCF model is updated; the water content data is detected through the updated RRCF model to obtain the abnormal score of each water content data point, so as to complete the water content monitoring of the oleic acid dehydration process for the preparation of diesel antiwear agent.

[0007] The beneficial effects are as follows: planning and determining abnormal performance according to the slope value of the curve can more accurately reflect the behavior of data points and their degree of abnormality, and improve the sensitivity and accuracy to data changes; by correcting the degree of abnormal performance and combining the difference between the reference value and the actual value, the anomalies in the data can be more accurately identified and corrected to avoid false alarms or omissions; updating the RRCF model according to the accuracy evaluation results can continuously improve the detection capability and accuracy of the model, and monitor and dynamically adjust the model in real time to make it more adaptable to changes in actual data; by accurately monitoring the water content during oleic acid dehydration, the key parameters in the production process can be effectively controlled to avoid quality problems caused by abnormal water content; at the same time, anomalies can be discovered and corrected in a timely manner to prevent possible production problems, reduce equipment failures and maintenance costs, and improve production efficiency.

[0008] Furthermore, the curve fitting is performed using the least square method, and during the curve fitting, the abscissa is the time sequence of the data points, and the ordinate is the actual value of the data points.

[0009] Furthermore, the planned slope value satisfies the following relationship:

[0010] ; In the formula, For the The slope value after planning for the data points, For the The slope value of the data point, is the upper limit of the preset planning range. It is the lower limit of the preset planning range.

[0011] The beneficial effect is that by planning the slope value, any slope value that exceeds the range will be corrected to the preset maximum or minimum value, which can avoid the adverse impact of extreme values ​​on subsequent analysis and ensure that the data is smoothly transitioned within a given range, thereby improving the stability of the model and prediction accuracy.

[0012] Furthermore, the abnormal performance degree satisfies the following relationship:

[0013] ; In the formula, The first The degree of abnormal performance of the data point, is the number of data points in the data sequence, is the slope value after planning for the jth data point in the data sequence, Indicates The number of data points within a preset local range of data points, For the The preset local range of the data point The slope value after planning for the data points.

[0014] The beneficial effect is that by quantifying the difference in slope value of each data point relative to the overall data sequence and the data points in its local range, the absolute difference between the global mean and the local mean is calculated, and the data points that deviate from the normal pattern in the local range are effectively identified, thereby detecting anomalies more accurately.

[0015] Furthermore, the modified abnormal performance degree satisfies the following relationship:

[0016] ; In the formula, The first The degree of abnormal performance after correction of the data point, The first The degree of abnormal performance of the data point, The first The reference value of the data point, The first The actual value of the data point, is a natural exponential function.

[0017] The beneficial effect is that the corrected abnormal performance degree can dynamically adjust the abnormal score by introducing the difference between the reference value and the actual value, and use the natural exponential function to quantify the difference between the reference value and the actual value, which can more accurately reflect the abnormal performance degree of each data point, improve the accuracy and stability of anomaly detection, and make the detection results more in line with the actual situation.

[0018] Furthermore, the reference value is a value obtained by interpolating the data point using cubic spline.

[0019] Furthermore, the accuracy satisfies the following relationship:

[0020] ; In the formula, is the accuracy of the RRCF model, is the number of data points in the data sequence, For the data points in the scatter plot to the regression curve The distance is a natural exponential function.

[0021] The beneficial effects are as follows: by evaluating the distance from the data point to the regression curve in the scatter plot, the prediction accuracy of the RRCF model is quantified. Through the negative relationship of the exponential function, the increase in distance will significantly reduce the accuracy value, which can effectively highlight the performance of the model when dealing with abnormal data; applying the exponential function after calculating the mean makes the accuracy evaluation smoother and more robust, reduces the excessive influence of a single data point on the overall accuracy score, and improves reliability.

[0022] Furthermore, updating the RRCF model according to the accuracy includes: in response to the accuracy being no greater than a preset threshold, determining that the detection efficiency of the RRCF model is abnormal, and updating the RRCF model.

[0023] Furthermore, the method for completing the water content monitoring of the oleic acid dehydration process for preparing the diesel antiwear agent includes: in response to the abnormal score value of the water content data point being greater than a preset abnormal threshold, determining that the water content in the oleic acid dehydration process is abnormal, and issuing an early warning prompt, thereby completing the water content monitoring of the oleic acid dehydration process for preparing the diesel antiwear agent.

[0024] The present invention has the following beneficial effects:

[0025] The accuracy of the RRCF model is continuously evaluated through the changing characteristics of the water content data, that is, the accuracy of the model is monitored, and then the RRCF model is updated according to the accuracy. This can more accurately determine when the model needs to be updated, avoid unnecessary updates that lead to a waste of computing resources, and better adapt to changes in data distribution, thereby improving the robustness of the model and thereby improving the detection efficiency of water content data in the oleic acid dehydration process. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0027] Figure 1 The present invention is a flowchart of the steps of a method for monitoring water content in a dehydration process of oleic acid for preparing a diesel antiwear agent according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0029] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0030] Because the RRCF algorithm needs to regularly update the RRCF model to adapt to the new data distribution when processing the time series data of water content in the process of dehydration with oleic acid, but the traditional regular update strategy may lead to unnecessary updates when the data distribution is stable, resulting in a waste of computing resources or failure to respond to rapid changes in data distribution in a timely manner. Therefore, the present invention will evaluate the accuracy of the RRCF model through the change characteristics of water content data, thereby improving the anomaly detection efficiency of the RRCF algorithm.

[0031] See also Figure 1 , which shows a flow chart of the steps of a method for monitoring water content in a dehydration process of oleic acid for preparing a diesel antiwear agent provided by an embodiment of the present invention, the method comprising the following steps:

[0032] S1: Data sequence of water content data collected in real time during the oleic acid dehydration process.

[0033] A microwave moisture sensor is used to collect real-time data series of water content in the dehydration process of oleic acid used in the preparation of diesel antiwear agent.

[0034] Implementers can set the acquisition frequency according to actual implementation conditions, for example, 10 Hz.

[0035] S2: performing curve fitting on each data point in the data sequence and each data point within a preset local range of each data point.

[0036] Specifically, the curve fitting is performed using the least square method, and during the curve fitting, the abscissa is the time sequence of the data points, and the ordinate is the actual value of the data points.

[0037] It should be noted that, because the slope of the function ranges from negative infinity to positive infinity, if the slope of a certain data point is too large, it will interfere with the overall change trend of the data sequence. Therefore, the present invention plans the slope value by limiting the upper and lower limits.

[0038] According to the slope value of each data point on the fitting curve, the slope value of each data point is planned.

[0039] Specifically, the planned slope value satisfies the following relationship:

[0040] ;

[0041] In the formula, For the The slope value after planning for the data points, For the The slope value of the data point, is the upper limit of the preset planning range. It is the lower limit of the preset planning range.

[0042] Implementers can set the upper and lower limits based on actual implementation conditions. For example, the upper limit is 1 and the lower limit is -1.

[0043] S3: Determine the degree of abnormal performance of each data point in the data sequence.

[0044] It should be noted that the degree of abnormal performance can be obtained by the difference between the overall change trend of the data and the local change trend. The change trend of the data can be expressed by the average slope of the fitting curve of the data range of the required measurement. The physical meaning of the average slope is: when the average slope is positive, the larger the average slope, the faster the water content increases; when the average slope is negative, the smaller the average slope, the faster the water content decreases. Then the data change trend within the local range of the data point can be obtained by calculating the average value of the slope values ​​of each data point after planning within the preset local range of the data point; and the change trend of the overall water content data can be obtained by calculating the average value of the slope values ​​of each data point of the data sequence after planning.

[0045] Specifically, the abnormal performance degree satisfies the following relationship:

[0046] ;

[0047] In the formula, The first The degree of abnormal performance of the data point, is the number of data points in the data sequence, is the slope value after planning for the jth data point in the data sequence, Indicates The number of data points within a preset local range of data points, For the The preset local range of the data point The slope value after planning for the data points.

[0048] S4: Correct the degree of abnormal performance by comparing the difference between the reference value and the actual value of each data point in the data sequence.

[0049] It should be noted that, during the process of collecting moisture content data, the microwave moisture sensor is susceptible to the influence of surrounding electromagnetic interference, which causes the microwave signal to be disturbed and then causes isolated noise points to appear in the moisture content data. The appearance of noise points will affect the calculation of the above-mentioned abnormal performance degree. Therefore, it is necessary to correct the abnormal performance degree through the isolation of the data points. The higher the isolation, the greater the noise interference to the data points, and the lower the credibility of the abnormal performance degree, and the lower the abnormal performance degree. The isolation of the data point can be represented by the difference between the actual value and the reference value of the data point. The larger the difference, the higher the isolation. Therefore, the abnormal performance degree is corrected through the difference between the reference value and the actual value.

[0050] Specifically, the modified abnormal performance degree satisfies the following relationship:

[0051] ;

[0052] In the formula, The first The degree of abnormal performance after correction of the data points, The first The degree of abnormal performance of the data point, The first The reference value of the data point, The first The actual value of the data point, is a natural exponential function.

[0053] Specifically, the reference value is a value obtained by interpolating the data point using cubic spline.

[0054] S5: Construct a scatter plot to determine the accuracy of the RRCF model.

[0055] The RRCF model is used to obtain the anomaly score of each data point in the data sequence. A scatter plot is constructed with the degree of abnormal performance of each data point as the ordinate and the anomaly score as the abscissa to determine the accuracy of the RRCF model.

[0056] It should be noted that the accuracy of the RRCF model can be expressed by the correlation between the degree of abnormal performance of the data points in the water content data series and the abnormal score. By establishing a coordinate system with the degree of abnormal performance as the vertical axis and the abnormal score as the horizontal axis, and constructing a scatter plot, the clustering of the data point distribution in the scatter plot is observed. The greater the clustering, the better the correlation; because the degree of abnormal performance and the abnormal score should be positively correlated, the clustering refers to the data points clustering near a curve (the curve shows the positive correlation between the two). Therefore, the least squares method can be used to find the polynomial regression curve of the scatter plot, and the average distance of all points to the regression curve can be found. The inverse of the average distance can represent the clustering, that is, the larger the average distance, the worse the clustering.

[0057] Specifically, the accuracy satisfies the following relationship:

[0058] ;

[0059] In the formula, is the accuracy of the RRCF model, is the number of data points in the data sequence, For the The regression curve of the data points in the scatter plot to the scatter plot The distance is a natural exponential function.

[0060] S6: Update the RRCF model according to the accuracy.

[0061] Specifically, updating the RRCF model according to the accuracy includes:

[0062] In response to the accuracy being no greater than a preset threshold, it is determined that the detection efficiency of the RRCF model is abnormal, and the RRCF model is updated.

[0063] Implementers can set the threshold and the interval time for obtaining the accuracy of the RRCF model according to the actual implementation situation. For example, the threshold is 0.65 and the interval time is 10s.

[0064] S7: Use the updated RRCF model to perform detection to complete the water content monitoring of the oleic acid dehydration process.

[0065] The moisture content data is tested using the updated RRCF model to obtain the abnormal score of each moisture content data point in order to complete the moisture content monitoring of the oleic acid dehydration process for the preparation of diesel antiwear agent.

[0066] Specifically, the water content monitoring of the oleic acid dehydration process for preparing diesel antiwear agent includes:

[0067] In response to the abnormal score value of the water content data point being greater than the preset abnormal threshold, it is determined that the water content in the oleic acid dehydration process is abnormal, and an early warning prompt is issued to complete the water content monitoring of the oleic acid dehydration process for the preparation of diesel antiwear agent.

[0068] Implementers can set the abnormal threshold according to actual implementation conditions, for example, 0.7.

[0069] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent, characterized in that: include: A data sequence for real-time acquisition of water content data during the oleic acid dehydration process; Perform curve fitting for each data point in the data sequence and each data point within a preset local range of each data point respectively, plan the slope value of each data point according to the slope value of each data point on the fitting curve, and the planned slope value is positively correlated with the slope value of the data point; determine the degree of abnormal performance of each data point in the data sequence, and the degree of abnormal performance is positively correlated with the difference between the mean of the slope values ​​of each data point in the data sequence after planning and the mean of the slope values ​​of each data point in the preset local range of the data point after planning; correct the degree of abnormal performance by the difference between the reference value and the actual value of each data point in the data sequence, and the corrected degree of abnormal performance is positively correlated with the degree of abnormal performance and negatively correlated with the difference; obtain the abnormal score of each data point in the data sequence through the RRCF model, construct a scatter plot with the abnormal performance degree of each data point as the ordinate and the abnormal score as the abscissa, and determine the accuracy of the RRCF model, and the accuracy is negatively correlated with the mean of the distance of each data point in the scatter plot to the regression curve of the scatter plot; update the RRCF model according to the accuracy; The degree of abnormal performance satisfies the following relationship: ; In the formula, The first The degree of abnormal performance of the data point, is the number of data points in the data sequence, is the slope value after planning for the jth data point in the data sequence, Indicates The number of data points within a preset local range of data points, For the The preset local range of the data point The slope value after planning for the data points; The moisture content data is tested using the updated RRCF model to obtain the abnormal score of each moisture content data point in order to complete the moisture content monitoring of the oleic acid dehydration process for the preparation of diesel antiwear agent.

2. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The curve fitting is performed using the least square method. During the curve fitting, the horizontal axis is the time sequence of the data points, and the vertical axis is the actual value of the data points.

3. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The planned slope value satisfies the following relationship: ； In the formula, For the The slope value after planning for the data points, For the The slope value of the data point, is the upper limit of the preset planning range. It is the lower limit of the preset planning range.

4. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The modified abnormal performance degree satisfies the following relationship: ； In the formula, The first The degree of abnormal performance after correction of the data points, The first The degree of abnormal performance of the data point, The first The reference value of the data point, The first The actual value of the data point, is a natural exponential function.

5. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1 or 4, characterized in that: The reference value is a value obtained by interpolating the data point using cubic spline.

6. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The accuracy satisfies the following relationship: ； In the formula, is the accuracy of the RRCF model, is the number of data points in the data sequence, For the The regression curve of the data points in the scatter plot to the scatter plot The distance is a natural exponential function.

7. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The updating of the RRCF model according to the accuracy includes: In response to the accuracy being no greater than a preset threshold, it is determined that the detection efficiency of the RRCF model is abnormal, and the RRCF model is updated.

8. The method for monitoring water content in the dehydration process of oleic acid for preparing diesel antiwear agent according to claim 1, characterized in that: The method for completing the water content monitoring of the oleic acid dehydration process for preparing diesel antiwear agent comprises: In response to the abnormal score value of the water content data point being greater than the preset abnormal threshold, it is determined that the water content in the oleic acid dehydration process is abnormal, and an early warning prompt is issued to complete the water content monitoring of the oleic acid dehydration process for the preparation of diesel antiwear agent.

Images