PDS ball valve state monitoring method and system based on multi-sensor fusion

Abstract

The invention relates to the field of ball valve state monitoring, in particular to a PDS ball valve state monitoring method and system based on multi-sensor fusion, and the method comprises the steps: obtaining ball valve operation data through a plurality of sensors, carrying out the preprocessing, fusing an initial matrix, carrying out the PCA decomposition, calculating the correlation between an abnormal feature sequence and a principal component score sequence, and carrying out the recognition of the abnormal feature sequence. And quantifying the abnormal sensitivity of each principal component by combining the average abnormal level of the abnormal feature sequence, calculating the importance degree of each principal component and sorting based on the abnormal sensitivity and combining the variance contribution rate of each principal component, and selecting reconstructed data with the optimal number of reserved principal components for denoising. And a reconstruction error is calculated for the denoised operation data, a local abnormal factor is calculated by improving an LOF algorithm, and the operation state of the ball valve is judged. According to the method, efficient denoising is achieved by optimizing principal component selection, the reliability of anomaly judgment is enhanced in combination with the LOF algorithm improved by reconstruction errors, and the real-time performance, the response speed and the stability of a monitoring system are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ball valve state monitoring. In particular, it relates to a PDS ball valve state monitoring method and system based on multi-sensor fusion. BACKGROUND

[0002] Ball valves are a kind of equipment used in industrial fields, and the performance and state of ball valves are crucial to the stability and safety of the entire production process. The importance of ball valve state monitoring lies in the ability to detect potential faults in a timely manner and avoid production interruptions and economic losses caused by equipment failures. Through effective state monitoring, maintenance plans can be arranged in advance, extending the service life of equipment, and improving production efficiency and product quality.

[0003] In existing PDS ball valve state monitoring methods, single sensors are usually used to collect data such as vibration data or pressure data to monitor the state of the ball valve and determine whether an anomaly has occurred. However, the industrial environment is complex and there is a lot of noise interference. These noises can cause deviations in the data collected by single sensors, making data analysis prone to misjudgment or missed judgment. Misjudgment refers to identifying normal operating conditions as abnormal conditions, leading to unnecessary maintenance work by enterprises, resulting in waste of resources and reduction of production efficiency. Missed judgment refers to the failure to detect potential fault risks in a timely manner, allowing faults to persist and possibly worsen, ultimately leading to more serious equipment damage. SUMMARY

[0004] To solve the problem of single sensor data collection in existing PDS ball valve state monitoring methods being easily affected by industrial environment noise interference, leading to misjudgment or missed judgment in data analysis, and causing unnecessary maintenance or more serious equipment damage and production accidents, the present application provides solutions in the following aspects.

[0005] In a first aspect, the PDS ball valve state monitoring method based on multi-sensor fusion comprises: obtaining and preprocessing the running data of the ball valve based on multiple sensors; fusing the initial matrix of the preprocessed running data and performing PCA decomposition to obtain the principal components and score sequences, calculating the correlation between the abnormal feature sequence of the running data sequence of each sensor and the score sequence of the principal components, quantifying the abnormal sensitivity of each principal component in combination with the average abnormal level of the abnormal feature sequence, and calculating the importance of each principal component based on the abnormal sensitivity and the variance contribution rate of each principal component; sorting the principal components according to the importance, traversing different numbers of retained principal components, calculating the denoising effect of different numbers of principal components, selecting the number of principal components corresponding to the maximum denoising effect, reconstructing the running data, and completing denoising; calculating the reconstruction error of each sampling point for the denoised running data, improving the distance between sampling points in the LOF algorithm in combination with the reconstruction error, obtaining the weighted distance, calculating the local anomaly factor of each sampling point based on the weighted distance, and judging the running state of the PDS ball valve according to the local anomaly factor.

[0006] Preferably, the abnormal feature sequence is calculated in the following manner: Wherein each row in the initial matrix represents a sample, and each column represents the data dimension of a sensor. For the running data sequence of each sensor in the initial matrix, the mean value of the running data sequence of each sensor is calculated, the ratio of each sampling point data to the mean value is calculated, the absolute value of 1 minus the ratio is taken as the relative deviation degree of each sampling point, the information entropy value of each sensor running data sequence is calculated, and the information entropy value is multiplied by the relative deviation degree to obtain the abnormal feature of each sampling point data, and the abnormal feature sequence of the running data sequence is constructed.

[0007] Preferably, the score sequence is obtained in the following steps: Calculate the mean value of each column of the initial matrix, subtract the mean value from each feature to obtain the decentered initial matrix, calculate the covariance initial matrix of the decentered initial matrix, perform eigenvalue decomposition on the covariance initial matrix to obtain the eigenvalues and corresponding eigenvectors, arrange all the eigenvectors into the initial matrix, and project the decentered initial matrix onto the eigenvector initial matrix to obtain the initial matrix of the principal components, wherein each row in the initial matrix of the principal components represents the score sequence of each principal component.

[0008] Preferably, the abnormal sensitivity of the principal component is calculated in the following manner: Take the score sequence of any principal component as the target sequence, calculate the Pearson correlation coefficient between the target sequence and each abnormal feature sequence, take the absolute value of the Pearson correlation coefficient, and weight the sum of the absolute values of the Pearson correlation coefficients in combination with the mean value of each abnormal feature, calculate the sum of the mean values of all abnormal features for normalization, and obtain the abnormal sensitivity of the principal component.

[0009] Preferably, the calculation method of the importance of each principal component comprises: The ratio of the eigenvalue of each principal component to the sum of all eigenvalues is taken as the variance contribution rate, and the product of the variance contribution rate and the anomaly sensitivity is taken as the importance of each principal component.

[0010] Preferably, the calculation method of the denoising effect of different principal component numbers comprises: The score sequence of the retained principal components is extracted from the initial matrix to construct a reconstruction matrix, the reconstruction error of each sampling point in the initial matrix and the reconstruction matrix is calculated, the ratio between the sum of anomaly sensitivities corresponding to the retained principal component numbers and the sum of anomaly sensitivities corresponding to all principal component numbers is calculated, the reconstruction error is taken as the influence weight, the reconstruction error is processed using a normalization function, and the denoising effect corresponding to the retained different principal component numbers is obtained.

[0011] Preferably, the acquisition method of the weighted distance comprises: The Euclidean distance between any two sampling points in the denoised running data is calculated, the reconstruction error of each sampling point is calculated respectively, the maximum value of the reconstruction error is selected, the ratio between the maximum value of the reconstruction error and the average value of the reconstruction errors of all sampling points is used to weight the Euclidean distance, and the improved weighted distance is obtained.

[0012] In a second aspect, a PDS ball valve state monitoring system based on multi-sensor fusion comprises a processor and a memory, and the memory stores computer program instructions.

[0013] The present application has the following effects: 1、The present application calculates the importance of each principal component and sorts them, selects the principal component number with the best denoising effect by traversing different retained principal component numbers, and reconstructs and denoises the running data. The present application can remove noise interference in the data, retain information most valuable for anomaly judgment, reduce the complexity and amount of data processing, improve the efficiency of data processing, complete the analysis and judgment of the running state of the ball valve faster, give accurate results in time, improve the real-time performance and response speed of the monitoring system, and better meet the real-time requirement of equipment state monitoring in industrial production.

[0014] 2、The application calculates the reconstruction error of each sampling point, and combines the reconstruction error to improve the distance between the sampling points in the LOF algorithm to obtain a weighted distance, and calculates the local anomaly factor of each sampling point based on the weighted distance, so as to judge the running state of the ball valve. The reliability of the abnormality judgment is enhanced, so that the monitoring system can more stably and accurately judge the running state of the ball valve when facing different types of faults and complex operating environments, reduces the misjudgment and omission caused by noise interference and other factors, and improves the adaptability and stability of the monitoring system. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a method flowchart of steps S1-S4 in the PDS ball valve state monitoring method based on multi-sensor fusion according to an embodiment of the application.

[0016] Figure 2 is a structural block diagram of the PDS ball valve state monitoring system based on multi-sensor fusion according to an embodiment of the application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments of the application.

[0018] Referring to Figure 1 The PDS ball valve state monitoring method based on multi-sensor fusion includes steps S1-S4, and specifically as follows: S1: Obtain the running data of the ball valve based on multiple sensors and perform preprocessing.

[0019] Multiple types of sensors are deployed at key positions of the PDS (Products Discharge System, powder discharge system) ball valve, including vibration sensors, acoustic emission (AE) sensors, pressure sensors, and temperature sensors, etc., to comprehensively monitor the running state of the ball valve.

[0020] The sensors collect vibration, acoustic emission, pressure, temperature, and other signal data in real time during the running process of the PDS ball valve according to the set collection interval (such as 1 second). The collection interval can be adjusted according to actual monitoring requirements and equipment running characteristics to ensure the timeliness and integrity of the data.

[0021] The collected data of different dimensions are accurately aligned according to their time stamps. Ensure that all data points have consistent time references in the time sequence, so as to compare and analyze at the same time point. For data with inconsistent time stamps, interpolation or other time alignment techniques are used to keep the data synchronized in the time dimension.

[0022] The data of each dimension is normalized respectively to eliminate the influence of different dimensions and ranges. Minimum-maximum normalization, Z-score normalization, etc. are adopted, and appropriate normalization strategy is selected according to the characteristics of data and analysis requirements.

[0023] In an industrial environment, the data collected by various sensors during the operation of a PDS ball valve often contains a large amount of noise, which can seriously interfere with the accurate identification of the state of the ball valve. In order to effectively remove noise, the existing technology usually adopts the PCA (Principal Component Analysis) decomposition and reconstruction method to process the data. However, the traditional PCA method mainly finds a set of projection directions that can maximize the variance after projection during the optimization process, and selects the principal components based on the variance contribution rate (eigenvalue), thereby retaining the direction with the largest variation in the data. Although this approach can capture the main information in the data, it has certain limitations. Since the performance characteristics of the data in each dimension are not completely the same, some low-variance directions may contain important abnormal information, such as weak impact signals or local leakage signals, etc. These are often ignored in the traditional PCA method, resulting in unsatisfactory noise removal and affecting the accuracy of subsequent anomaly detection.

[0024] In order to overcome the shortcomings of the traditional PCA method, the abnormal features of the running data in the principal components are further introduced based on the sorting of the principal components based on variance, in order to enhance the sensitivity to fault signals and improve the noise removal effect and the accuracy of subsequent anomaly detection. Before implementing this scheme, the abnormal feature degree of the running data needs to be quantified. The specific steps are as follows: S2: fuse the initial matrix of the preprocessed running data and perform PCA decomposition to obtain the principal components and score sequence, calculate the correlation between the abnormal feature sequence of the running data sequence of each sensor and the score sequence of the principal components, combine the average abnormal level of the abnormal feature sequence, quantify the abnormal sensitivity of each principal component, and calculate the importance of each principal component based on the abnormal sensitivity and the variance contribution rate of each principal component.

[0025] Wherein, each row in the initial matrix represents a sample, and each column represents a data dimension of a sensor. For the running data sequence of each sensor in the initial matrix, the mean value of the running data sequence of each sensor is calculated, the ratio of each sampling point data to the mean value is calculated, the absolute value of 1 minus the ratio is taken as the relative deviation degree of each sampling point, the information entropy value of each sensor running data sequence is calculated, and the information entropy value is multiplied by the relative deviation degree to obtain the abnormal feature of each sampling point data, and the abnormal feature sequence of the running data sequence is constructed.

[0026] Specifically, the abnormal feature sequence satisfies the following relationship: ; In the formula, represents the abnormal feature of the data of the i th sampling point collected by the i th sensor, represents the information entropy value of the running data sequence collected by the i th sensor, The information entropy value can reflect the complexity of the data distribution in the running data sequence, and the greater the information entropy value, the greater the fluctuation degree of the running data in the PDS ball valve running process, and the greater the possibility of abnormality, represents the data of the i th sampling point collected by the i th sensor in the initial matrix represents the mean value of all data in the running data sequence of the i th sensor in the initial matrix The closer the ratio is to 1, the closer the data of the i th sampling point collected by the i th sensor is to the average level, and the smaller the possibility of abnormality; on the contrary, the farther the ratio is from 1, the farther the data of the i th sampling point collected by the i th sensor is from the average level, and the greater the possibility of abnormality. By introducing information entropy and mean value ratio to quantify the abnormality degree, the data fluctuation and deviation in the PDS ball valve running process can be effectively captured. Not only the complexity and volatility of the data are considered, but also the deviation degree of each sampling point from the overall average level is measured through the mean value ratio. Specifically, the greater the information entropy value, the greater the fluctuation degree of the data sequence, and the higher the possibility of abnormality; and the farther the mean value ratio is from 1, the more abnormal the sampling point data is. By combining the deviation degrees of information entropy and mean value ratio, the present application can more accurately identify potential fault signals, thereby significantly enhancing the sensitivity of PCA to fault signals. Unlike traditional PCA which only focuses on the direction with large energy, by quantifying abnormal features, the loss of key abnormal information is avoided, and thus the denoising effect and the accuracy of subsequent anomaly detection are effectively improved.

[0027] The score sequence is obtained by the following steps:

[0028] The score sequence is obtained by the following steps: ​​​​​​​​​​​​Calculate the mean of each column of the initial matrix, subtract the mean from each feature to obtain the decentralized initial matrix, calculate the covariance initial matrix of the decentralized initial matrix, perform eigenvalue decomposition on the covariance initial matrix to obtain eigenvalues ​​and corresponding eigenvectors, arrange all eigenvectors into the initial matrix, and project the decentralized initial matrix onto the eigenvector initial matrix to obtain the initial matrix of principal components, where each column of the initial matrix of principal components represents the score sequence of each principal component.

[0029] Using the score sequence of any principal component as the target sequence, calculate the Pearson correlation coefficient between the target sequence and each abnormal feature sequence. Take the absolute value of the Pearson correlation coefficient and combine it with the mean of each abnormal feature to take the absolute value of the Pearson correlation coefficient for weighted summation. Calculate the sum of the means of all abnormal features for normalization processing to obtain the abnormality sensitivity of the principal component.

[0030] Specifically, the anomaly sensitivity of the principal components satisfies the following relationship: ; In the formula, Indicates the first Abnormal sensitivity of principal components Indicates abnormal characteristics The total number, Indicates the first The score sequence of each principal component With the A sequence of abnormal features The Pearson correlation coefficient is used to quantify the correlation between two factors, and the range of values ​​for the Pearson correlation coefficient is... Here, the Pearson correlation coefficient is used to characterize the correlation without considering whether it is positive or negative. Therefore, the absolute value of the Pearson correlation coefficient is used for calculation. The closer the absolute value is to 1, the stronger the correlation, and the closer the absolute value is to 0, the weaker the correlation. Indicates the first One abnormal feature The mean.

[0031] In other words, the anomaly sensitivity of each principal component is quantified by multiplying the correlation strength between the anomaly feature sequence of each sensor and the score sequence of the principal component by the average anomaly feature of the corresponding sensor, and then summing the results over all sensors. This comprehensively considers the correlation between the principal components and the anomaly features of each sensor, as well as the average anomaly level of each sensor.

[0032] Simply considering the correlation between principal components and sensor data may not be sufficient to fully reflect the sensitivity of principal components to anomalies, because some sensors may also exhibit high anomalous characteristics under normal conditions, but these characteristics may not be significant. By multiplying by the average anomalous characteristics, the contribution of each sensor to the anomaly sensitivity of the principal components can be measured more accurately.

[0033] If a sensor's anomalous feature sequence is highly correlated with principal components but has a low average anomalous feature, then the sensor's contribution to principal component anomalous sensitivity may be small. Conversely, if a sensor's anomalous feature sequence is low correlated with principal components but has a high average anomalous feature, then the sensor may still make a significant contribution to principal component anomalous sensitivity.

[0034] By calculating the anomaly sensitivity of each principal component and combining it with variance (eigenvalues) to determine the importance of the principal components, the priority of the principal components can be dynamically adjusted. This approach considers not only the variance contribution of the principal component score sequences but also the sensitivity of the principal component score sequences to anomaly features, thus more effectively retaining principal components related to anomalies and improving the robustness and accuracy of the monitoring system.

[0035] For example, in industrial environments, the score sequences of certain low-variance principal components may contain important anomalous information (such as weak shock signals or localized leak signals), which traditional PCA methods might overlook. By introducing anomaly sensitivity, it can be ensured that these important low-variance principal components are preserved when reconstructing the data.

[0036] By quantifying the anomaly sensitivity of principal components, principal components related to anomalies can be identified more accurately, thereby retaining more anomaly information and reducing misjudgments and omissions during data reconstruction and denoising.

[0037] The methods for calculating the importance of each principal component include: The proportion of each principal component's eigenvalue to the sum of all eigenvalues ​​is used as the variance contribution rate, and the product of the variance contribution rate and the anomaly sensitivity is used as the importance of each principal component.

[0038] Specifically, the importance level satisfies the following relationship: ; In the formula, Indicates the first The importance of each principal component Indicates the first The anomaly sensitivity of each principal component is considered. The greater the anomaly sensitivity, the more important the principal component is, and the higher its ranking. It should not be discarded when reconstructing and denoising data. Indicates the first The standard PCA algorithm sorts the principal components based on their eigenvalues. The eigenvalues ​​represent the variance in each eigenvector direction. The larger the eigenvalue, the more information is captured in that direction. Represents the sum and ratio of all eigenvalues. The larger the value, the more likely it is to indicate the first... The larger the proportion of the eigenvalues ​​of the principal components, the better the reflection of the first principal component. The greater the variance captured by each principal component, the higher its importance.

[0039] By combining anomaly sensitivity with variance, the priority of principal components can be dynamically adjusted, thereby prioritizing the retention of information closely related to the abnormal state of the PDS ball valve. This significantly improves the system's adaptability to non-stationary signals and effectively reduces the adverse effects of noise interference on the state identification process. This method can significantly improve the accuracy of the entire monitoring system, enhance the reliability of fault diagnosis, and ensure accurate monitoring and timely early warning of the PDS ball valve's state.

[0040] S3: Sort the principal components according to their importance, traverse different numbers of retained principal components, calculate the denoising effect of different numbers of principal components, select the number of principal components corresponding to the largest denoising effect, reconstruct the running data, and complete the denoising.

[0041] Extract the score sequence of retained principal components from the initial matrix, construct the reconstruction matrix, calculate the reconstruction error of each sampling point in the initial matrix and the reconstruction matrix, calculate the ratio between the sum of anomaly sensitivities corresponding to the number of retained principal components and the sum of anomaly sensitivities corresponding to the number of all principal components, use the reconstruction error as the influence weight, and process the reconstruction error using a normalization function to obtain the denoising effect corresponding to different numbers of retained principal components.

[0042] Specifically, the noise reduction effect satisfies the following relationship: ; In the formula, Indicates to retain the previous text Denoising effect of principal components Indicates to retain the previous text The reconstruction error between the denoised running data and the running data is a quantitative indicator of the amount of information lost, reflecting the denoising effect. The larger the reconstruction error, the more information is discarded, the more noise is removed, and the better the denoising effect. Represents the normalization function. Indicates to retain the previous text Number of principal components Indicates the first Abnormal sensitivity of principal components Indicates to retain the previous text The sum of the anomaly sensitivities of the principal components, represents the number of all principal components, represents the sum of the anomaly sensitivities of all principal components, and the first The greater the ratio of the number of principal components to the anomaly sensitivity, the more details related to the anomaly are retained in the energy, and the better the denoising effect.

[0043] For example, assume that there is a 4-dimensional data set containing 100 samples. After PCA processing, 4 principal components are obtained, and the anomaly sensitivities are , respectively. .

[0044] The principal components are sorted according to the eigenvalues: , and the corresponding eigenvalues are ; Retain 1 principal component: the reconstruction matrix is constructed by , and the reconstruction error and denoising effect are calculated .

[0045] Retain 2 principal components: the reconstruction matrix is constructed by and , and the reconstruction error and denoising effect are calculated .

[0046] Retain 3 principal components: the reconstruction matrix is constructed by , and , and the reconstruction error and denoising effect are calculated .

[0047] Retain 4 principal components: the reconstruction matrix is constructed by , , and , and the reconstruction error and denoising effect are calculated .

[0048] The reconstruction error reflects the reconstruction quality of the data after retaining principal components, and the ratio of the anomaly sensitivity reflects the proportion of the retained principal components in the total anomaly sensitivity. By comparing under different numbers, the optimal number of retained principal components can be found to achieve the best denoising effect.

[0049] S4: Calculate the reconstruction error of each sampling point for the denoised running data, combine the reconstruction error to improve the distance between sampling points in the LOF algorithm, obtain the weighted distance, calculate the local anomaly factor of each sampling point based on the weighted distance, and judge the running state of the PDS ball valve according to the local anomaly factor.

[0050] Calculate the Euclidean distance between any two sampling points in the denoised running data, calculate the reconstruction error for each sampling point, select the maximum value of the reconstruction error, and weight the Euclidean distance by the ratio between the maximum value of the reconstruction error and the mean of the reconstruction errors of all sampling points to obtain the improved weighted distance.

[0051] Specifically, the weighted distance satisfies the following relationship: ; In the formula, This represents the weighted distance between two sampling points. This represents the Euclidean distance between two sampling points. Indicates the first Reconstruction error of data collected from each sampling point Indicates the first Reconstruction error of data collected from each sampling point This represents the mean of the reconstruction error across all sampling points.

[0052] In other words, reconstruction error weighting amplifies the distance between sampling points containing uncaptured anomalous features (such as weak fault signals) and other normal points, thereby improving the sensitivity and isolation of the LOF (Local Outlier Factor) algorithm for outliers. By introducing reconstruction error weighting, the impact of the reconstruction process on the calculation of anomaly severity is reduced. This enables more accurate identification and differentiation between weak fault signals and normal signals, avoiding misjudgments caused by information loss during reconstruction, and improving the accuracy and robustness of anomaly detection.

[0053] Based on the improved weighted distance, the local anomaly factor of each sampling point is calculated using the LOF algorithm. When the local anomaly factor is greater than or equal to 1, it is determined that the PDS ball valve has an abnormal condition at the time point corresponding to the sampling point. Conversely, when it is less than 1, it is determined that the PDS ball valve is in normal operation. The threshold in this embodiment is set to 1, which can be adjusted according to specific circumstances.

[0054] This invention also provides a PDS ball valve condition monitoring system based on multi-sensor fusion. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the PDS ball valve status monitoring method based on multi-sensor fusion according to the first aspect of the present invention. The system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, the settings and functions of which are known in the art and will not be described further here.

[0055] It should be noted that, for the person skilled in the art, several modifications and improvements can be made without departing from the inventive concept, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application patent should be subject to the appended claims.

Claims

1. A PDS ball valve condition monitoring method based on multi-sensor fusion, characterized in that, include: The ball valve's operating data is acquired using multiple sensors and then preprocessed. The preprocessed running data is fused with the initial matrix and PCA decomposition is performed to obtain principal components and score sequences. The correlation between the abnormal feature sequences of the running data sequences of each sensor and the score sequences of the principal components is calculated. The abnormality sensitivity of each principal component is quantified by combining the average abnormality level of the abnormal feature sequences. The importance of each principal component is calculated based on the abnormality sensitivity and the variance contribution rate of each principal component. The principal components are sorted according to their importance, different numbers of retained principal components are traversed, and the denoising effect of different numbers of principal components is calculated. The number of principal components corresponding to the largest denoising effect is selected, and the running data is reconstructed to complete the denoising. For the denoised running data, the reconstruction error of each sampling point is calculated. The distance between sampling points in the LOF algorithm is improved by combining the reconstruction error to obtain the weighted distance. The local anomaly factor of each sampling point is calculated based on the weighted distance, and the operating status of the PDS ball valve is determined based on the local anomaly factor.

2. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The calculation method of the abnormal feature sequence, include: In the initial matrix, each row represents a sample and each column represents the data dimension of a sensor. For the operating data sequence of each sensor in the initial matrix, the mean of the operating data sequence of each sensor is calculated, the ratio of each sampling point data to the mean is calculated, the absolute value of 1 minus the ratio is taken as the relative deviation of each sampling point, the information entropy value of the operating data sequence of each sensor is calculated, and the information entropy value is multiplied by the relative deviation to obtain the abnormal features of the data of each sampling point, thus constructing the abnormal feature sequence of the operating data sequence.

3. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The method for obtaining the score sequence includes the following steps: Calculate the mean of each column of the initial matrix, subtract the mean from each feature to obtain the decentralized initial matrix, calculate the covariance initial matrix of the decentralized initial matrix, perform eigenvalue decomposition on the covariance initial matrix to obtain eigenvalues ​​and corresponding eigenvectors, arrange all eigenvectors into the initial matrix, and project the decentralized initial matrix onto the eigenvector initial matrix to obtain the initial matrix of principal components, where each column of the initial matrix of principal components represents the score sequence of each principal component.

4. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The calculation method for the abnormal sensitivity of the principal component includes: Using the score sequence of any principal component as the target sequence, calculate the Pearson correlation coefficient between the target sequence and each abnormal feature sequence. Take the absolute value of the Pearson correlation coefficient and combine it with the mean of each abnormal feature to take the absolute value of the Pearson correlation coefficient for weighted summation. Calculate the sum of the means of all abnormal features for normalization processing to obtain the abnormality sensitivity of the principal component.

5. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The methods for calculating the importance of each principal component include: The proportion of each principal component's eigenvalue to the sum of all eigenvalues ​​is used as the variance contribution rate, and the product of the variance contribution rate and the anomaly sensitivity is used as the importance of each principal component.

6. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The calculation method for the denoising effect of different principal component numbers includes: Extract the score sequence of retained principal components from the initial matrix, construct the reconstruction matrix, calculate the reconstruction error of each sampling point in the initial matrix and the reconstruction matrix, calculate the ratio between the sum of anomaly sensitivities corresponding to the number of retained principal components and the sum of anomaly sensitivities corresponding to the number of all principal components, use the reconstruction error as the influence weight, and process the reconstruction error using a normalization function to obtain the denoising effect corresponding to different numbers of retained principal components.

7. The PDS ball valve condition monitoring method based on multi-sensor fusion according to claim 1, characterized in that, The weighted distance is obtained in the following ways: Calculate the Euclidean distance between any two sampling points in the denoised running data, calculate the reconstruction error for each sampling point, select the maximum value of the reconstruction error, and weight the Euclidean distance by the ratio between the maximum value of the reconstruction error and the mean of the reconstruction errors of all sampling points to obtain the improved weighted distance.

8. A PDS ball valve condition monitoring system based on multi-sensor fusion, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the PDS ball valve status monitoring method based on multi-sensor fusion according to any one of claims 1-7.

Images