CN117090989A

CN117090989A - Electric gate valve with monitoring system

Info

Publication number: CN117090989A
Application number: CN202211656082.6A
Authority: CN
Inventors: 康康; 张昌挺; 杨光标; 蔡楠楠; 郑特雷; 冉银波; 陈坚
Original assignee: ZHEJIANG DECA CONTROL VALVE METER CO Ltd
Current assignee: ZHEJIANG DECA CONTROL VALVE METER CO Ltd
Priority date: 2022-12-22
Filing date: 2022-12-22
Publication date: 2023-11-21

Abstract

The application discloses an electric gate valve with a monitoring system, which adopts an artificial intelligent detection technology based on deep learning to extract implicit relevance characteristic distribution information of each current transient waveform through sampling analysis on a plurality of current transient waveforms in the working process of the electric gate valve, converts the implicit relevance characteristic distribution information into digital signals based on an analog-to-digital conversion idea, then extracts multi-scale relevance characteristic distribution information among the current digital signals, and carries out classification processing on whether the electric gate valve is normal or not through a classifier based on the fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, and whether the electric valve has faults or not can be accurately judged.

Description

Electric gate valve with monitoring system

Technical Field

The present application relates to the field of electric gate valves, and more particularly to an electric gate valve with a monitoring system.

Background

The electric valve is controlled by an electric actuator, so that the valve is opened and closed. The action force distance is larger than that of the electric valve, the action speed of the electric gate valve can be adjusted, the structure is simple, the maintenance is easy, the damage caused by clamping is not easy due to the buffer characteristic of gas in the action process, but an air source is needed, and the control system is also more complex than that of the electric valve. The electric gate valve has sensitive response, safety and reliability, and a plurality of manufacturers with high control requirements are specially provided with a compressed air station for the pneumatic instrument control element. Electric is the device which needs electricity and can control the flow rate.

In the working process of the electric gate valve, although the electric gate valve has the advantages of sensitive response, safety and reliability, when a medium flows through the valve body, a vortex phenomenon is generated in the upper cavity, so that radial vibration of the valve core is caused, cavitation phenomenon is generated, and flushing is caused on valve internal parts, thereby influencing the service life of the valve and the regulation performance of the valve. And when the valve encounters impact and 'water hammer', the whole transmission mechanical mechanism can generate fatigue, and the service life of the valve is influenced.

Accordingly, an electrically powered gate valve with a monitoring system is desired.

Disclosure of Invention

The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides an electric gate valve with a monitoring system, which adopts an artificial intelligent detection technology based on deep learning to extract implicit relevance characteristic distribution information of each current instantaneous waveform by sampling and analyzing a plurality of current instantaneous waveforms in the working process of the electric gate valve, converts the implicit relevance characteristic distribution information into digital signals based on an analog-to-digital conversion idea, then extracts multi-scale relevance characteristic distribution information among the current digital signals, and carries out classification processing on whether the performance of the electric gate valve is normal or not through a classifier based on the fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, and whether the electric valve has faults or not can be accurately judged.

Accordingly, according to one aspect of the present application, there is provided an electrically powered gate valve having a monitoring system, comprising: the monitoring data acquisition module is used for acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; the analog-to-digital conversion module is used for carrying out analog-to-digital conversion on the current instantaneous waveform analog quantity signals so as to obtain a plurality of current digital signals; the analog signal graph feature extraction module is used for respectively passing the plurality of current instantaneous waveform analog quantity signals through a first convolution neural network model serving as a filter to obtain a plurality of current analog feature vectors; the analog signal feature association coding module is used for performing two-dimensional arrangement on the plurality of current analog feature vectors to obtain a feature matrix, and then obtaining a current analog association feature matrix through a second convolution neural network model serving as a feature extractor; the digital signal feature extraction module is used for arranging the plurality of current digital signals into input vectors and then obtaining current digital association feature vectors through the multi-scale neighborhood feature extraction module; the characteristic enhancement module is used for carrying out characteristic data enhancement on the current digital association characteristic vector by using a Gaussian density chart so as to obtain a current digital association characteristic matrix; the multi-mode feature fusion module is used for fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; the class center offset correction module is used for carrying out class center offset correction on the classification characteristic matrix to obtain a corrected classification characteristic matrix; and the monitoring result generation module is used for enabling the corrected classification characteristic matrix to pass through a classifier to obtain a classification result, and the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

In the above electric gate valve with a monitoring system, the analog signal pattern feature extraction module is further configured to: each layer using the first convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the first convolutional neural network model is the current simulation feature vector, and the input of the first layer of the first convolutional neural network model is the current instantaneous waveform analog quantity signal.

In the electric gate valve with the monitoring system, the analog signal characteristic association coding module is further used for: each layer using the second convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the second convolutional neural network model is the current simulation correlation feature matrix, and the input of the first layer of the second convolutional neural network model is the feature matrix.

In the electric gate valve with the monitoring system, the multi-scale neighborhood feature extraction module comprises a first convolution layer, a second convolution layer and a cascade layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first scale, the second convolution layer uses a one-dimensional convolution layer with a second scale, and the first scale is different from the second scale.

In the above electric gate valve with a monitoring system, the digital signal feature extraction module includes: the first scale feature extraction unit is used for carrying out one-dimensional convolution encoding on the input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a first scale current digital association feature vector; wherein, the formula is:

wherein a is the width of a first convolution kernel in the X direction, F (a) is a first convolution kernel parameter vector, G (X-a) is a local vector matrix operated with a convolution kernel function, w is the size of the first convolution kernel, X represents the input vector, and Cov (X) represents the first scale current digital association feature vector; the second scale feature extraction unit is used for carrying out one-dimensional convolution encoding on the input vector by using a second convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a second scale current digital association feature vector; wherein, the formula is:

Wherein b is the width of the second convolution kernel in the X direction, F (b) is a second convolution kernel parameter vector, G (X-b) is a local vector matrix operated with a convolution kernel function, m is the size of the second convolution kernel, X represents the input vector, and Cov (X) represents the second scale current digital correlation feature vector; and the multi-scale fusion unit is used for cascading the first scale current digital association feature vector and the second scale current digital association feature vector by using a cascading layer of the multi-scale neighborhood feature extraction module so as to obtain the temperature difference feature vector.

In the above electric gate valve with a monitoring system, the feature enhancement module includes: the Gaussian density map construction unit is used for constructing a Gaussian density map of the current digital correlation feature vector, wherein the mean value vector of the Gaussian density map is the current digital correlation feature vector, and the value of each position in the covariance matrix of the Gaussian density map is the variance between the feature values of two corresponding positions in the current digital correlation feature vector; and the discretization unit is used for carrying out Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map so as to obtain a current digital association characteristic matrix.

In the above electric gate valve with a monitoring system, the multi-modal feature fusion module is further configured to: fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix according to the following formula; wherein, the formula is:

wherein M is _c For the classification feature matrix, M _a For the current digital correlation characteristic matrix, M _b The correlation feature matrix is simulated for the current,and (c) representing element addition at corresponding positions of the current digital correlation feature matrix and the current analog correlation feature matrix, wherein alpha and beta are weighting parameters for controlling balance between the current digital correlation feature matrix and the current analog correlation feature matrix in the classification feature matrix.

In the above electric gate valve with a monitoring system, the center-like offset correction module includes: the diagonal matrix conversion unit is used for respectively converting the current analog correlation characteristic matrix and the current digital correlation characteristic matrix into diagonal matrices to obtain a first diagonal matrix and a second diagonal matrix; the class node topology-class center fusion unit is used for carrying out class node topology-class center fusion on the first focusing matrix and the second diagonal matrix to obtain a fusion feature matrix; and the mapping unit is used for multiplying the fusion feature matrix and the classification feature matrix by a matrix to obtain the corrected classification feature matrix.

In the electric gate valve with the monitoring system, the topology-class center fusion unit of the class node is further configured to: performing node-like topology-like center fusion on the first diagonal matrix and the second diagonal matrix by using the following formula to obtain the fusion feature matrix; wherein, the formula is:

wherein M is ₁ 、M ₂ And M _c The first diagonal matrix, the second diagonal matrix and the fused feature matrix,and ∈H.alpha.expressed as Kronecker product and Hadamard product, D (M) ₁ ,M ₂ ) For a position-by-position distance matrix between the first diagonal matrix and the second diagonal matrix, exp (·) represents an exponential operation of the matrix, which represents calculating a natural exponential function value that is a power of a eigenvalue of each position in the matrix.

In the above electric gate valve with a monitoring system, the monitoring result generating module includes: the unfolding unit is used for unfolding the corrected classification characteristic matrix into a classification characteristic vector according to a row vector or a column vector; the full-connection coding unit is used for carrying out full-connection coding on the classification characteristic vectors by using a full-connection layer of the classifier so as to obtain coded classification characteristic vectors; and the classification result generating unit is used for inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

According to another aspect of the present application, there is also provided a method of operating an electric gate valve having a monitoring system, comprising: acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; analog-to-digital conversion is carried out on the current instantaneous waveform analog quantity signals to obtain a plurality of current digital signals; respectively passing the current instantaneous waveform analog quantity signals through a first convolution neural network model serving as a filter to obtain a plurality of current simulation feature vectors; two-dimensionally arranging the plurality of current simulation feature vectors to form a feature matrix, and then obtaining a current simulation correlation feature matrix through a second convolution neural network model serving as a feature extractor; the plurality of current digital signals are arranged into input vectors and then pass through a multi-scale neighborhood feature extraction module to obtain current digital association feature vectors; performing feature data enhancement on the current digital association feature vector by using a Gaussian density chart to obtain a current digital association feature matrix; fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; performing class center offset correction on the classification feature matrix to obtain a corrected classification feature matrix; and passing the corrected classification characteristic matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

In the above method for operating an electric gate valve with a monitoring system, the step of passing the plurality of current transient waveform analog quantity signals through a first convolutional neural network model as a filter to obtain a plurality of current analog feature vectors includes: each layer using the first convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the first convolutional neural network model is the current simulation feature vector, and the input of the first layer of the first convolutional neural network model is the current instantaneous waveform analog quantity signal.

In the above method for operating an electric gate valve with a monitoring system, the two-dimensionally arranging the plurality of current analog feature vectors as a feature matrix, and then obtaining a current analog correlation feature matrix by using a second convolutional neural network model as a feature extractor, includes: each layer using the second convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the second convolutional neural network model is the current simulation correlation feature matrix, and the input of the first layer of the second convolutional neural network model is the feature matrix.

In the operation method of the electric gate valve with the monitoring system, the multi-scale neighborhood feature extraction module comprises a first convolution layer, a second convolution layer and a cascade layer connected with the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first scale, the second convolution layer uses a one-dimensional convolution layer with a second scale, and the first scale is different from the second scale.

In the above operation method of the electric gate valve with a monitoring system, the step of obtaining the current digital correlation feature vector by the multi-scale neighborhood feature extraction module after arranging the plurality of current digital signals as the input vector includes: performing one-dimensional convolution encoding on the input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a first-scale current digital association feature vector; wherein, the formula is:

wherein a is the width of a first convolution kernel in the X direction, F (a) is a first convolution kernel parameter vector, G (X-a) is a local vector matrix operated with a convolution kernel function, w is the size of the first convolution kernel, X represents the input vector, and Cov (X) represents the first scale current digital association feature vector; performing one-dimensional convolution encoding on the input vector by using a second convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a second-scale current digital association feature vector; wherein, the formula is:

Wherein b is the width of the second convolution kernel in the X direction, F (b) is a second convolution kernel parameter vector, G (X-b) is a local vector matrix operated with a convolution kernel function, m is the size of the second convolution kernel, X represents the input vector, and Cov (X) represents the second scale current digital correlation feature vector; and cascading the first scale current digital correlation feature vector and the second scale current digital correlation feature vector by using a cascading layer of the multi-scale neighborhood feature extraction module to obtain the temperature difference feature vector.

In the above operation method of the electric gate valve with a monitoring system, the performing feature level data enhancement on the current digital correlation feature vector by using a gaussian density chart to obtain a current digital correlation feature matrix includes: constructing a Gaussian density map of the current digital correlation feature vector, wherein the mean value vector of the Gaussian density map is the current digital correlation feature vector, and the value of each position in a covariance matrix of the Gaussian density map is the variance between the feature values of two corresponding positions in the current digital correlation feature vector; and performing Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain a current digital association characteristic matrix.

In the above operation method of the electric gate valve with a monitoring system, the fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix includes: fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix according to the following formula; wherein, the formula is:

In the above operation method of the electric gate valve with a monitoring system, the performing the class center offset correction on the classification feature matrix to obtain a corrected classification feature matrix includes: converting the current analog correlation characteristic matrix and the current digital correlation characteristic matrix into diagonal matrixes respectively to obtain a first diagonal matrix and a second diagonal matrix; performing node-like topology-like center fusion on the first focusing matrix and the second diagonal matrix to obtain a fusion feature matrix; and multiplying the fusion feature matrix and the classification feature matrix by a matrix to obtain the corrected classification feature matrix.

In the above operation method of the electric gate valve with the monitoring system, the performing node-like topology-like center fusion on the first focusing matrix and the second diagonal matrix to obtain a fused feature matrix includes: performing node-like topology-like center fusion on the first diagonal matrix and the second diagonal matrix by using the following formula to obtain the fusion feature matrix; wherein, the formula is:

wherein M is ₁ 、M ₂ And M _c The first diagonal matrix, the second diagonal matrix and the fused feature matrix,and ∈H.alpha.expressed as Kronecker product and Hadamard product, D (M) ₁ ,M ₂ ) An exp (·) table for a position-by-position distance matrix between the first diagonal matrix and the second diagonal matrixAn exponential operation of a matrix is shown, which represents the calculation of a natural exponential function value that is a power of a eigenvalue of each position in the matrix.

In the above operation method of the electric gate valve with the monitoring system, the passing the corrected classification feature matrix through a classifier to obtain a classification result, where the classification result is used to indicate whether the performance of the electric gate valve is normal, includes: expanding the corrected classification characteristic matrix into classification characteristic vectors according to row vectors or column vectors; performing full-connection coding on the classification feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

Compared with the prior art, the electric gate valve with the monitoring system provided by the application adopts an artificial intelligent detection technology based on deep learning, so that the implicit relevance characteristic distribution information of each current instantaneous waveform is extracted by sampling and analyzing a plurality of current instantaneous waveforms in the working process of the electric gate valve, the multi-scale relevance characteristic distribution information among all current digital signals is extracted later based on the conversion thought of analog to digital signals, and the classification processing of whether the performance of the electric gate valve is normal or not is carried out by a classifier based on the fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, and whether the electric valve has faults or not can be accurately judged.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.

Fig. 1 is an application scenario diagram of an electric gate valve with a monitoring system according to an embodiment of the present application.

Fig. 2 is a block diagram of an electrically powered gate valve with a monitoring system in accordance with an embodiment of the application.

Fig. 3 is a schematic diagram of an electric gate valve with a monitoring system according to an embodiment of the application.

Fig. 4 is a block diagram of a center-like offset correction module in an electric gate valve with a monitoring system in accordance with an embodiment of the present application.

Fig. 5 is a flow chart of a method of operating an electric gate valve with a monitoring system according to an embodiment of the application.

Detailed Description

Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.

Summary of the application

As described above, in the working process of the electric gate valve, although the electric gate valve has the advantages of sensitive response, safety and reliability, when a medium flows through the valve body, a vortex phenomenon is generated in the upper cavity, so that radial oscillation of the valve core is caused, cavitation phenomenon is generated, and flushing is caused on valve internal parts, thereby influencing the service life of the valve and the regulation performance of the valve. And when the valve encounters impact and 'water hammer', the whole transmission mechanical mechanism can generate fatigue, and the service life of the valve is influenced. Accordingly, an electrically powered gate valve with a monitoring system is desired.

Accordingly, considering that input and output of current are required to control the gate valve in the working process of the electric gate valve, the performance of the electric gate valve can be detected on line in real time based on the collected current signals by sampling a plurality of current instantaneous waveforms in the working process of the electric gate valve so as to determine whether the electric gate valve has faults or not. However, since the instantaneous waveforms of the plurality of currents are small-scale signals, the signals are easily interfered by external factors in the process of acquisition, so that noise exists in the acquired signals, and the detection accuracy is affected. In the detection process, as the correlation exists among the current transient signals, the difficulty is how to dig out the correlation relation among the current transient signals, and meanwhile, how to perform real-time online detection on the performance of the electric valve based on the current signals so as to judge whether the electric valve has faults.

In recent years, deep learning and neural networks have been widely used in the fields of computer vision, natural language processing, text signal processing, and the like. In addition, deep learning and neural networks have also shown levels approaching and even exceeding humans in the fields of image classification, object detection, semantic segmentation, text translation, and the like.

The development of deep learning and neural networks provides new solutions and schemes for mining the association relation among the current transient signals and carrying out real-time on-line detection on the performance of the electric valve based on the current signals.

Specifically, in the technical scheme of the application, an artificial intelligent detection technology based on deep learning is adopted to sample and analyze a plurality of current transient waveforms in the working process of the electric gate valve, so that implicit relevance characteristic distribution information of each current transient waveform is extracted, multiscale relevance characteristic distribution information among each current digital signal is extracted based on an analog-to-digital conversion idea, and classification processing of whether the electric gate valve is normal in performance or not is performed through a classifier based on fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, so that whether the electric valve has a fault or not can be accurately judged, and the normal operation of the electric gate valve is ensured.

More specifically, in the technical scheme of the application, firstly, a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve are obtained. Then, since the plurality of current instantaneous waveform analog quantity signals in the electric gate valve operation process are waveform images in nature, feature mining of the plurality of current instantaneous waveform analog quantity signals is performed by using a first convolution neural network model which is a filter and has excellent expression in terms of implicit feature extraction of images, so that implicit feature distribution information of each current instantaneous waveform analog quantity signal in the plurality of current instantaneous waveform analog quantity signals is extracted respectively, and a plurality of current analog feature vectors are obtained.

Further, considering that there are correlation relations of different degrees between the current instantaneous waveform analog quantity signals, in order to sufficiently mine out the correlation information to perform performance detection of the electric gate valve, the plurality of current instantaneous waveform analog quantity signals are further two-dimensionally arranged into a characteristic matrix and then processed in a second convolution neural network model serving as a characteristic extractor to extract hidden correlation characteristics between the current instantaneous waveform analog quantity signals, so that a current analog correlation characteristic matrix is obtained.

Then, when the performance of the electric gate valve is detected by using the plurality of current instantaneous waveform analog quantity signals in the working process of the electric gate valve, the accuracy of detection is reduced due to the fact that the plurality of current instantaneous waveform analog quantity signals are easily interfered by external noise in the collecting process, so that the accuracy of fault diagnosis is affected. Therefore, in the technical scheme of the application, the performance detection is comprehensively carried out by selecting the fusion characteristics of the plurality of current instantaneous waveform analog quantity signals and the digital signals in the high-dimensional space based on the action process of the electric gate valve, and the detection accuracy can be obviously improved. Specifically, first, analog-to-digital conversion is performed on the plurality of current instantaneous waveform analog signals to obtain a plurality of current digital signals. Then, considering that the plurality of current digital signals have different degrees of correlation, the plurality of current digital signals are further arranged into input vectors and then subjected to feature mining through a multi-scale neighborhood feature extraction module, so that multi-scale correlation features among the plurality of current digital signals under different digital signal scales are extracted, and a current digital correlation feature vector is obtained.

Next, in order to improve the accuracy of the extraction of the hidden correlation features for the respective current digital signals, data enhancement is required for the hidden correlation features of the respective current digital signals in a high-dimensional feature space, considering that the respective current digital signals have volatility and uncertainty. It should be understood that, as a learning target of the neural network model, the gaussian density map may represent a joint distribution in the case where a plurality of feature values constitute an overall distribution due to probability density thereof, that is, a feature distribution is taken as a priori distribution, and probability density under the influence of correlation with other a priori distribution positions at each a priori distribution position is obtained as a posterior distribution, thereby describing the feature distribution more accurately in a higher dimension. Therefore, in the technical scheme of the application, the prior distribution, namely the Gaussian distribution, of each current digital signal can be used for carrying out data enhancement on the implicit association features of each current digital signal, namely, the feature data enhancement is carried out on the current digital association feature vector by using a Gaussian density chart so as to obtain a current digital association feature matrix. Specifically, a Gaussian density map of the current digital correlation feature vector is constructed, and Gaussian discretization processing is performed on the Gaussian density map so that information loss is not generated when data features are amplified, and thus a current digital correlation feature matrix is obtained.

Further, the current digital correlation characteristic matrix and the current analog correlation characteristic matrix are fused to represent the analog-digital fusion characteristic information of the correlation characteristic distribution of a plurality of current instantaneous waveform signals in a high-dimensional space in the working process of the electric gate valve, and the analog-digital fusion characteristic information is used as a classification characteristic matrix to be subjected to classification processing in a classifier, so that a classification result used for representing whether the performance of the electric gate valve is normal is obtained. Therefore, the performance of the electric valve can be detected in real time, so that whether the electric valve has faults or not can be accurately judged.

Particularly, in the technical scheme of the application, when the current digital correlation feature matrix and the current analog correlation feature matrix are fused to obtain the classification feature matrix, the current analog correlation feature matrix expresses correlation features among image semantics of a plurality of current instantaneous waveform analog quantity signals, and the current digital correlation feature matrix expresses multi-scale context correlation features of the plurality of current digital signals, so that the classification feature matrix may have class center offset, thereby influencing the accuracy of classification results of the classification feature matrix.

The current analog correlation matrix and the current digital correlation matrix are therefore preferably first converted into diagonal matrices, for example denoted as M ₁ And M ₂ Then, topology-class center fusion of class nodes is performed, expressed as:

and ∈H.alpha.expressed as Kronecker product and Hadamard product, D (M) ₁ ,M ₂ ) Is a feature matrix M ₁ And M ₂ A position-by-position distance matrix between, i.e. +.>

The applicant of the present application considers that in the classification problem of the classifier, if the current is modeled to be correlated with the feature matrix M ₁ And the current digital correlation characteristic matrix M ₂ The fused class nodes are expressed as tree forms, and then the respective class nodes are distributed as subtrees based on root nodes, so that the node distribution of the fused class nodes can be expressed as a sub-graph structure centered on the respective nodes based on the graph topology by utilizing the graph topology related between the nodes, thereby expressing the current simulation association characteristic matrix M ₁ And the current digital correlation characteristic matrix M ₂ The respective class nodes are in a subtree structure of the root to realize the current simulation association characteristic matrix M ₁ And the current digital correlation characteristic matrix M ₂ Is based on node-center-like topology fusion, thereby eliminating the current simulation correlation feature matrix M ₁ And the current digital correlation characteristic matrix M ₂ Center-of-class offset between.

Then, the fused feature matrix M _c And the classification characteristic matrix M _c Matrix multiplication is performed to obtain the classification characteristic matrix M _c Mapping into a fusion feature space for eliminating class center offset to promote the classification feature matrix M _c And (3) the accuracy of the classification result of the model is obtained, so that a corrected classification characteristic matrix is obtained. And then, classifying the corrected classification characteristic matrix through a classifier, so as to obtain a more accurate classification result which indicates whether the performance of the electric gate valve is normal. Therefore, the performance of the electric valve can be accurately detected in real time, so that whether the electric valve has a fault or not can be accurately judged, and the normal operation of the electric gate valve is ensured.

Based on this, the present application provides an electric gate valve with a monitoring system, comprising: the monitoring data acquisition module is used for acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; the analog-to-digital conversion module is used for carrying out analog-to-digital conversion on the current instantaneous waveform analog quantity signals so as to obtain a plurality of current digital signals; the analog signal graph feature extraction module is used for respectively passing the plurality of current instantaneous waveform analog quantity signals through a first convolution neural network model serving as a filter to obtain a plurality of current analog feature vectors; the analog signal feature association coding module is used for performing two-dimensional arrangement on the plurality of current analog feature vectors to obtain a feature matrix, and then obtaining a current analog association feature matrix through a second convolution neural network model serving as a feature extractor; the digital signal feature extraction module is used for arranging the plurality of current digital signals into input vectors and then obtaining current digital association feature vectors through the multi-scale neighborhood feature extraction module; the characteristic enhancement module is used for carrying out characteristic data enhancement on the current digital association characteristic vector by using a Gaussian density chart so as to obtain a current digital association characteristic matrix; the multi-mode feature fusion module is used for fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; the class center offset correction module is used for carrying out class center offset correction on the classification characteristic matrix to obtain a corrected classification characteristic matrix; and the monitoring result generation module is used for enabling the corrected classification characteristic matrix to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

Fig. 1 is an application scenario diagram of an electric gate valve with a monitoring system according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, a plurality of current transient waveform analog quantity signals during the action of an electric gate valve (e.g., V as illustrated in fig. 1) are acquired by an oscilloscope (e.g., O as illustrated in fig. 1). Further, the plurality of current instantaneous waveform analog quantity signals during the operation of the electric gate valve are input into a data processor (for example, P as illustrated in fig. 1) of the electric gate valve with a monitoring system, wherein the data processor can process the plurality of current instantaneous waveform analog quantity signals during the operation of the electric gate valve based on a predetermined algorithm to obtain a classification result for indicating whether the performance of the electric gate valve is normal.

Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.

Exemplary System

Fig. 2 is a block diagram of an electrically powered gate valve with a monitoring system in accordance with an embodiment of the application. As shown in fig. 2, the electric gate valve 100 with a monitoring system according to an embodiment of the present application includes: the monitoring data acquisition module 110 is used for acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; the analog-to-digital conversion module 120 is configured to perform analog-to-digital conversion on the plurality of current instantaneous waveform analog signals to obtain a plurality of current digital signals; the analog signal pattern feature extraction module 130 is configured to pass the plurality of current transient waveform analog signals through a first convolutional neural network model serving as a filter to obtain a plurality of current analog feature vectors; the analog signal feature association encoding module 140 is configured to two-dimensionally arrange the plurality of current analog feature vectors into a feature matrix, and then obtain a current analog association feature matrix through a second convolutional neural network model serving as a feature extractor; the digital signal feature extraction module 150 is configured to arrange the plurality of current digital signals into an input vector, and then pass through the multi-scale neighborhood feature extraction module to obtain a current digital correlation feature vector; the feature enhancement module 160 is configured to perform feature level data enhancement on the current digital correlation feature vector by using a gaussian density chart to obtain a current digital correlation feature matrix; the multi-mode feature fusion module 170 is configured to fuse the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; the class center offset correction module 180 is configured to perform class center offset correction on the classification feature matrix to obtain a corrected classification feature matrix; and a monitoring result generating module 190, configured to pass the corrected classification feature matrix through a classifier to obtain a classification result, where the classification result is used to indicate whether the performance of the electric gate valve is normal.

Fig. 3 is a schematic diagram of an electric gate valve with a monitoring system according to an embodiment of the application. As shown in fig. 3, firstly, obtaining a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; then, carrying out analog-to-digital conversion on the current instantaneous waveform analog quantity signals to obtain a plurality of current digital signals; then, the current instantaneous waveform analog quantity signals are respectively passed through a first convolution neural network model serving as a filter to obtain a plurality of current analog feature vectors; then, the plurality of current analog feature vectors are arranged in two dimensions to form a feature matrix, then the feature matrix is obtained through a second convolution neural network model serving as a feature extractor, and meanwhile, the plurality of current digital signals are arranged to form an input vector, and then the input vector is passed through a multi-scale neighborhood feature extraction module to obtain a current digital association feature vector; then, carrying out feature data enhancement on the current digital association feature vector by using a Gaussian density chart to obtain a current digital association feature matrix; then, fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; performing class center offset correction on the classification characteristic matrix to obtain a corrected classification characteristic matrix; and finally, the corrected classification characteristic matrix passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

In the electric gate valve 100 with the monitoring system, the monitoring data acquisition module 110 is configured to acquire a plurality of current instantaneous waveform analog signals during the action of the electric gate valve. As described above, in the working process of the electric gate valve, although the electric gate valve has the advantages of sensitive response, safety and reliability, when a medium flows through the valve body, a vortex phenomenon is generated in the upper cavity, so that radial oscillation of the valve core is caused, cavitation phenomenon is generated, and flushing is caused on valve internal parts, thereby influencing the service life of the valve and the regulation performance of the valve. And when the valve encounters impact and 'water hammer', the whole transmission mechanical mechanism can generate fatigue, and the service life of the valve is influenced. Accordingly, an electrically powered gate valve with a monitoring system is desired.

Specifically, in the technical scheme of the application, an artificial intelligent detection technology based on deep learning is adopted to sample and analyze a plurality of current transient waveforms in the working process of the electric gate valve, so that implicit relevance characteristic distribution information of each current transient waveform is extracted, multiscale relevance characteristic distribution information among each current digital signal is extracted based on an analog-to-digital conversion idea, and classification processing of whether the electric gate valve is normal in performance or not is performed through a classifier based on fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, so that whether the electric valve has a fault or not can be accurately judged, and the normal operation of the electric gate valve is ensured. More specifically, in the technical scheme of the application, firstly, a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve are obtained. Wherein the plurality of current transient waveform analog signals may be acquired by an oscilloscope.

In the electric gate valve 100 with the monitoring system, the analog-to-digital conversion module 120 is configured to perform analog-to-digital conversion on the plurality of current instantaneous waveform analog signals to obtain a plurality of current digital signals. Analog-to-digital conversion, also referred to herein as analog-to-digital conversion (a/D), refers to the conversion of an analog signal to a digital signal, which is the process of converting a continuously varying analog quantity into a set of discrete digital points. Analog-to-digital conversion mainly includes two processes, sampling and quantization.

In the electric gate valve 100 with the monitoring system, the analog signal pattern feature extraction module 130 is configured to pass the plurality of current transient waveform analog signals through a first convolutional neural network model as a filter to obtain a plurality of current analog feature vectors. Since the plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve are waveform images in nature, the characteristic mining of the plurality of current instantaneous waveform analog quantity signals is carried out by adopting a first convolution neural network model which is a filter and has excellent expression in the aspect of implicit characteristic extraction of images so as to extract implicit characteristic distribution information of each current instantaneous waveform analog quantity signal in the plurality of current instantaneous waveform analog quantity signals respectively, thereby obtaining a plurality of current analog characteristic vectors.

Specifically, in the embodiment of the present application, the analog signal graphic feature extraction module 130 is further configured to: each layer using the first convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the first convolutional neural network model is the current simulation feature vector, and the input of the first layer of the first convolutional neural network model is the current instantaneous waveform analog quantity signal.

In the electric gate valve 100 with the monitoring system, the analog signal feature correlation encoding module 140 is configured to two-dimensionally arrange the plurality of current analog feature vectors into a feature matrix, and then obtain a current analog correlation feature matrix by using a second convolutional neural network model as a feature extractor. In view of the fact that there are correlation relations of different degrees between the current instantaneous waveform analog quantity signals, in order to fully mine out the correlation information to detect the performance of the electric gate valve, the current instantaneous waveform analog quantity signals are further processed in a second convolution neural network model serving as a feature extractor after being two-dimensionally arranged into a feature matrix, so that hidden correlation features between the current instantaneous waveform analog quantity signals are extracted, and a current analog correlation feature matrix is obtained.

Specifically, in the embodiment of the present application, the analog signal feature association encoding module 140 is further configured to: each layer using the second convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the second convolutional neural network model is the current simulation correlation feature matrix, and the input of the first layer of the second convolutional neural network model is the feature matrix.

In the electric gate valve 100 with the monitoring system, the digital signal feature extraction module 150 is configured to arrange the plurality of current digital signals into input vectors and then obtain current digital correlation feature vectors through the multi-scale neighborhood feature extraction module. When the electric gate valve performance detection is carried out by using the plurality of current instantaneous waveform analog quantity signals in the working process of the electric gate valve, the detection accuracy is reduced due to the fact that the plurality of current instantaneous waveform analog quantity signals are easily interfered by external noise in the collecting process, and therefore the fault diagnosis accuracy is affected. Therefore, in the technical scheme of the application, the performance detection is comprehensively carried out by selecting the fusion characteristics of the plurality of current instantaneous waveform analog quantity signals and the digital signals in the high-dimensional space based on the action process of the electric gate valve, and the detection accuracy can be obviously improved. Specifically, after analog-to-digital conversion is performed on the current instantaneous waveform analog signals to obtain a plurality of current digital signals, considering that the current digital signals have different degrees of correlation, the current digital signals are further arranged into input vectors, and then feature mining is performed in a multi-scale neighborhood feature extraction module, so that multi-scale correlation features among the current digital signals in different digital signal scales are extracted, and a current digital correlation feature vector is obtained.

In an embodiment of the present application, the multi-scale neighborhood feature extraction module includes a first convolution layer and a second convolution layer in parallel, and a cascade layer connected to the first convolution layer and the second convolution layer, where the first convolution layer uses a one-dimensional convolution kernel having a first scale, and the second convolution layer uses a one-dimensional convolution layer having a second scale, and the first scale is different from the second scale.

Specifically, in the embodiment of the present application, the digital signal feature extraction module 150 includes: the first scale feature extraction unit is used for carrying out one-dimensional convolution encoding on the input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a first scale current digital association feature vector; wherein, the formula is:

In the electric gate valve 100 with the monitoring system, the feature enhancement module 160 is configured to perform feature level data enhancement on the current digital correlation feature vector by using a gaussian density chart to obtain a current digital correlation feature matrix. In view of the volatility and uncertainty of the respective current digital signals, data enhancement is required for the implicit correlated features of the respective current digital signals in a high-dimensional feature space in order to improve the accuracy of the extraction of the hidden correlated features for the respective current digital signals. It should be understood that, as a learning target of the neural network model, the gaussian density map may represent a joint distribution in the case where a plurality of feature values constitute an overall distribution due to probability density thereof, that is, a feature distribution is taken as a priori distribution, and probability density under the influence of correlation with other a priori distribution positions at each a priori distribution position is obtained as a posterior distribution, thereby describing the feature distribution more accurately in a higher dimension. Therefore, in the technical scheme of the application, the prior distribution, namely the Gaussian distribution, of each current digital signal can be used for carrying out data enhancement on the implicit association features of each current digital signal, namely, the feature data enhancement is carried out on the current digital association feature vector by using a Gaussian density chart so as to obtain a current digital association feature matrix. Specifically, a Gaussian density map of the current digital correlation feature vector is constructed, and Gaussian discretization processing is performed on the Gaussian density map so that information loss is not generated when data features are amplified, and thus a current digital correlation feature matrix is obtained.

More specifically, in the embodiment of the present application, the feature enhancement module 160 first constructs, by a gaussian density map construction unit, a gaussian density map of the current digital correlation feature vector, where a mean vector of the gaussian density map is the current digital correlation feature vector, and a value of each position in a covariance matrix of the gaussian density map is a variance between feature values of two corresponding positions in the current digital correlation feature vector; and then, carrying out Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map by a discretization unit to obtain a current digital association characteristic matrix.

In the electric gate valve 100 with the monitoring system, the multi-mode feature fusion module 170 is configured to fuse the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix. The current digital correlation characteristic matrix and the current analog correlation characteristic matrix are fused to represent the analog-digital fusion characteristic information of the correlation characteristic distribution of a plurality of current instantaneous waveform signals in a high-dimensional space in the working process of the electric gate valve, and the analog-digital fusion characteristic information is used as a classification characteristic matrix.

Specifically, in the embodiment of the present application, the multi-modal feature fusion module 170 is further configured to: fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix according to the following formula; wherein, the formula is:

In the electric gate valve 100 with the monitoring system, the class center offset correction module 180 is configured to perform class center offset correction on the classification feature matrix to obtain a corrected classification feature matrix. In particular, in the technical scheme of the application, when the current digital correlation feature matrix and the current analog correlation feature matrix are fused to obtain the classification feature matrix, the current analog correlation feature matrix expresses correlation features among image semantics of a plurality of current instantaneous waveform analog quantity signals The current digital correlation feature matrix expresses multi-scale context correlation features of the plurality of current digital signals, so that class center offset can exist, thereby affecting the accuracy of classification results of the classification feature matrix. The current analog correlation matrix and the current digital correlation matrix are therefore preferably first converted into diagonal matrices, for example denoted as M ₁ And M ₂ And then carrying out topology-class center fusion of class nodes.

Fig. 4 is a block diagram of a center-like offset correction module in an electric gate valve with a monitoring system in accordance with an embodiment of the present application. As shown in fig. 4, the center-like offset correction module 180 includes: the diagonal matrix conversion unit 181 is configured to convert the current analog correlation characteristic matrix and the current digital correlation characteristic matrix into diagonal matrices respectively to obtain a first diagonal matrix and a second diagonal matrix; a class node topology-class center fusion unit 182, configured to perform class node topology-class center fusion on the first focusing matrix and the second diagonal matrix to obtain a fusion feature matrix; and a mapping unit 183, configured to perform matrix multiplication on the fusion feature matrix and the classification feature matrix to obtain the corrected classification feature matrix.

Specifically, in the embodiment of the present application, the topology-class center fusion unit 182 of the class node is further configured to: performing node-like topology-like center fusion on the first diagonal matrix and the second diagonal matrix by using the following formula to obtain the fusion feature matrix; wherein, the formula is:

wherein M is ₁ 、M ₂ And M _c The first diagonal matrix, the second diagonal matrix and the fused feature matrix,and ≡indicates Kronecker and Hadam of the matrix, respectivelyard product, D (M ₁ ,M ₂ ) For a position-by-position distance matrix between the first diagonal matrix and the second diagonal matrix, exp (·) represents an exponential operation of the matrix, which represents calculating a natural exponential function value that is a power of a eigenvalue of each position in the matrix.

Then, the fused feature matrix M _c And the classification characteristic matrix M _c Matrix multiplication is performed to obtain the classification characteristic matrix M _c Mapping into a fusion feature space for eliminating class center offset to promote the classification feature matrix M _c And (3) the accuracy of the classification result of the model is obtained, so that a corrected classification characteristic matrix is obtained.

In the electric gate valve 100 with the monitoring system, the monitoring result generating module 190 is configured to pass the corrected classification feature matrix through a classifier to obtain a classification result, where the classification result is used to indicate whether the performance of the electric gate valve is normal. Namely, the class boundary of the corrected classification characteristic matrix is divided and determined by the classifier, so that the classification result is obtained. Therefore, the performance of the electric valve can be accurately detected in real time, so that whether the electric valve has a fault or not can be accurately judged, and the normal operation of the electric gate valve is ensured.

Specifically, in the embodiment of the present application, the monitoring result generating module 190 includes: the unfolding unit is used for unfolding the corrected classification characteristic matrix into a classification characteristic vector according to a row vector or a column vector; the full-connection coding unit is used for carrying out full-connection coding on the classification characteristic vectors by using a full-connection layer of the classifier so as to obtain coded classification characteristic vectors; and the classification result generating unit is used for inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

In summary, the electric gate valve 100 with the monitoring system according to the embodiment of the application is illustrated, which adopts an artificial intelligent detection technology based on deep learning to sample and analyze a plurality of current transient waveforms in the working process of the electric gate valve, so as to extract implicit relevance characteristic distribution information of each current transient waveform, convert the implicit relevance characteristic distribution information into digital signals based on an analog-digital conversion idea, then extract multi-scale relevance characteristic distribution information among the current digital signals, and perform classification processing on whether the performance of the electric gate valve is normal or not based on the fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, and whether the electric valve has faults or not can be accurately judged.

Exemplary method

Fig. 5 is a flow chart of a method of operating an electric gate valve with a monitoring system according to an embodiment of the application. As shown in fig. 5, a method for operating an electric gate valve with a monitoring system according to an embodiment of the present application includes: s110, acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; s120, carrying out analog-to-digital conversion on the current instantaneous waveform analog quantity signals to obtain a plurality of current digital signals; s130, respectively passing the plurality of current instantaneous waveform analog quantity signals through a first convolution neural network model serving as a filter to obtain a plurality of current simulation feature vectors; s140, two-dimensionally arranging the plurality of current simulation feature vectors into a feature matrix, and then obtaining a current simulation association feature matrix through a second convolution neural network model serving as a feature extractor; s150, arranging the plurality of current digital signals into input vectors, and then obtaining current digital association feature vectors through a multi-scale neighborhood feature extraction module; s160, carrying out feature data enhancement on the current digital association feature vector by using a Gaussian density chart to obtain a current digital association feature matrix; s170, fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; s180, performing class center offset correction on the classification characteristic matrix to obtain a corrected classification characteristic matrix; and S190, passing the corrected classification characteristic matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

In one example, in the method for operating an electric gate valve with a monitoring system, the passing the plurality of current transient waveform analog quantity signals through a first convolutional neural network model as a filter to obtain a plurality of current analog eigenvectors includes: each layer using the first convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the first convolutional neural network model is the current simulation feature vector, and the input of the first layer of the first convolutional neural network model is the current instantaneous waveform analog quantity signal.

In an example, in the method for operating an electric gate valve with a monitoring system, the two-dimensionally arranging the plurality of current analog feature vectors into a feature matrix and then obtaining a current analog correlation feature matrix by using a second convolutional neural network model as a feature extractor includes: each layer using the second convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the second convolutional neural network model is the current simulation correlation feature matrix, and the input of the first layer of the second convolutional neural network model is the feature matrix.

In one example, in the method for operating an electric gate valve with a monitoring system, the multi-scale neighborhood feature extraction module includes a first convolution layer and a second convolution layer in parallel, and a cascade layer connected to the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first scale, and the second convolution layer uses a one-dimensional convolution layer with a second scale, and the first scale is different from the second scale.

In an example, in the method for operating an electric gate valve with a monitoring system, the step of arranging the plurality of current digital signals into input vectors and then obtaining current digital correlation feature vectors through a multi-scale neighborhood feature extraction module includes: performing one-dimensional convolution encoding on the input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a first-scale current digital association feature vector; wherein, the formula is:

In one example, in the method for operating an electric gate valve with a monitoring system, the performing feature level data enhancement on the current digital correlation feature vector by using a gaussian density chart to obtain a current digital correlation feature matrix includes: constructing a Gaussian density map of the current digital correlation feature vector, wherein the mean value vector of the Gaussian density map is the current digital correlation feature vector, and the value of each position in a covariance matrix of the Gaussian density map is the variance between the feature values of two corresponding positions in the current digital correlation feature vector; and performing Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map to obtain a current digital association characteristic matrix.

In one example, in the method for operating an electric gate valve with a monitoring system, the fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix includes: fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix according to the following formula; wherein, the formula is:

wherein M is _c For the classification featureMatrix, M _a For the current digital correlation characteristic matrix, M _b The correlation feature matrix is simulated for the current,and (c) representing element addition at corresponding positions of the current digital correlation feature matrix and the current analog correlation feature matrix, wherein alpha and beta are weighting parameters for controlling balance between the current digital correlation feature matrix and the current analog correlation feature matrix in the classification feature matrix.

In one example, in the method for operating an electric gate valve with a monitoring system, the performing a class center offset correction on the classification feature matrix to obtain a corrected classification feature matrix includes: converting the current analog correlation characteristic matrix and the current digital correlation characteristic matrix into diagonal matrixes respectively to obtain a first diagonal matrix and a second diagonal matrix; performing node-like topology-like center fusion on the first focusing matrix and the second diagonal matrix to obtain a fusion feature matrix; and multiplying the fusion feature matrix and the classification feature matrix by a matrix to obtain the corrected classification feature matrix.

In an example, in the method for operating an electric gate valve with a monitoring system, the performing node-like topology-like center fusion on the first focusing matrix and the second diagonal matrix to obtain a fused feature matrix includes: performing node-like topology-like center fusion on the first diagonal matrix and the second diagonal matrix by using the following formula to obtain the fusion feature matrix; wherein, the formula is:

In one example, in the operation method of the electric gate valve with the monitoring system, the step of passing the corrected classification feature matrix through a classifier to obtain a classification result, where the classification result is used to indicate whether the electric gate valve is normal, includes: expanding the corrected classification characteristic matrix into classification characteristic vectors according to row vectors or column vectors; performing full-connection coding on the classification feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

In summary, the operation method of the electric gate valve with the monitoring system of the embodiment of the application is clarified, which adopts an artificial intelligent detection technology based on deep learning to sample and analyze a plurality of current transient waveforms in the working process of the electric gate valve, so as to extract implicit relevance characteristic distribution information of each current transient waveform, convert the implicit relevance characteristic distribution information into digital signals based on an analog-to-digital conversion idea, then extract multi-scale relevance characteristic distribution information among each current digital signal, and classify whether the performance of the electric gate valve is normal or not by a classifier based on the fusion characteristics of the two. Therefore, the performance of the electric valve can be accurately detected in real time, and whether the electric valve has faults or not can be accurately judged.

The basic principles of the present application have been described above in connection with specific embodiments, however, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are merely examples and not intended to be limiting, and these advantages, benefits, effects, etc. are not to be considered as essential to the various embodiments of the present application. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, as the application is not necessarily limited to practice with the above described specific details.

The block diagrams of the devices, apparatuses, devices, systems referred to in the present application are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be made in the manner shown in the block diagrams. As will be appreciated by one of skill in the art, the devices, apparatuses, devices, systems may be connected, arranged, configured in any manner. Words such as "including," "comprising," "having," and the like are words of openness and mean "including but not limited to," and are used interchangeably therewith. The terms "or" and "as used herein refer to and are used interchangeably with the term" and/or "unless the context clearly indicates otherwise. The term "such as" as used herein refers to, and is used interchangeably with, the phrase "such as, but not limited to.

It is also noted that in the apparatus, devices and methods of the present application, the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present application.

The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit embodiments of the application to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, a person of ordinary skill in the art will recognize certain variations, modifications, alterations, additions, and subcombinations thereof.

Claims

1. An electrically powered gate valve with a monitoring system, comprising: the monitoring data acquisition module is used for acquiring a plurality of current instantaneous waveform analog quantity signals in the action process of the electric gate valve; the analog-to-digital conversion module is used for carrying out analog-to-digital conversion on the current instantaneous waveform analog quantity signals so as to obtain a plurality of current digital signals; the analog signal graph feature extraction module is used for respectively passing the plurality of current instantaneous waveform analog quantity signals through a first convolution neural network model serving as a filter to obtain a plurality of current analog feature vectors; the analog signal feature association coding module is used for performing two-dimensional arrangement on the plurality of current analog feature vectors to obtain a feature matrix, and then obtaining a current analog association feature matrix through a second convolution neural network model serving as a feature extractor; the digital signal feature extraction module is used for arranging the plurality of current digital signals into input vectors and then obtaining current digital association feature vectors through the multi-scale neighborhood feature extraction module; the characteristic enhancement module is used for carrying out characteristic data enhancement on the current digital association characteristic vector by using a Gaussian density chart so as to obtain a current digital association characteristic matrix; the multi-mode feature fusion module is used for fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix; the class center offset correction module is used for carrying out class center offset correction on the classification characteristic matrix to obtain a corrected classification characteristic matrix; and the monitoring result generation module is used for enabling the corrected classification characteristic matrix to pass through a classifier to obtain a classification result, and the classification result is used for indicating whether the performance of the electric gate valve is normal or not.

2. The electric gate valve with monitoring system of claim 1, wherein the analog signal pattern feature extraction module is further to: each layer using the first convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; non-linear activation is carried out on the pooled feature map so as to obtain an activated feature map; the output of the last layer of the first convolutional neural network model is the current simulation feature vector, and the input of the first layer of the first convolutional neural network model is the current instantaneous waveform analog quantity signal.

3. The electric gate valve with monitoring system of claim 2, wherein the analog signal feature-associated encoding module is further configured to: each layer using the second convolutional neural network model is performed in forward pass of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; carrying out mean pooling based on a local feature matrix on the convolution feature map to obtain a pooled feature map; non-linear activation is carried out on the pooled feature map so as to obtain an activated feature map; the output of the last layer of the second convolutional neural network model is the current simulation correlation feature matrix, and the input of the first layer of the second convolutional neural network model is the feature matrix.

4. The motorized gate valve with monitoring system of claim 3, wherein the multi-scale neighborhood feature extraction module comprises first and second parallel convolution layers, and a cascade layer connected to the first and second convolution layers, wherein the first convolution layer uses a one-dimensional convolution kernel having a first scale and the second convolution layer uses a one-dimensional convolution layer having a second scale, the first scale being different from the second scale.

5. The electric gate valve with monitoring system of claim 4, wherein the digital signal feature extraction module comprises: the first scale feature extraction unit is used for carrying out one-dimensional convolution encoding on the input vector by using a first convolution layer of the multi-scale neighborhood feature extraction module according to the following formula to obtain a first scale current digital association feature vector; wherein, the formula is:

Wherein b is the width of the second convolution kernel in the X direction, F (b) is a second convolution kernel parameter vector, G (X-b) is a local vector matrix operated with a convolution kernel function, m is the size of the second convolution kernel, X represents the input vector, and Cov (X) represents the second scale current digital correlation feature vector; and

and the multi-scale fusion unit is used for cascading the first scale current digital association feature vector and the second scale current digital association feature vector by using a cascading layer of the multi-scale neighborhood feature extraction module so as to obtain the temperature difference feature vector.

6. The electric gate valve with monitoring system of claim 5, wherein the feature enhancement module comprises: the Gaussian density map construction unit is used for constructing a Gaussian density map of the current digital correlation feature vector, wherein the mean value vector of the Gaussian density map is the current digital correlation feature vector, and the value of each position in the covariance matrix of the Gaussian density map is the variance between the feature values of two corresponding positions in the current digital correlation feature vector; and the discretization unit is used for carrying out Gaussian discretization on the Gaussian distribution of each position in the Gaussian density map so as to obtain a current digital association characteristic matrix.

7. The electric gate valve with monitoring system of claim 6, wherein the multi-modal feature fusion module is further to: fusing the current digital correlation feature matrix and the current analog correlation feature matrix to obtain a classification feature matrix according to the following formula; wherein, the formula is:

8. The electric gate valve with monitoring system of claim 7, wherein the center-like offset correction module comprises: the diagonal matrix conversion unit is used for respectively converting the current analog correlation characteristic matrix and the current digital correlation characteristic matrix into diagonal matrices to obtain a first diagonal matrix and a second diagonal matrix; the class node topology-class center fusion unit is used for carrying out class node topology-class center fusion on the first focusing matrix and the second diagonal matrix to obtain a fusion feature matrix; and the mapping unit is used for multiplying the fusion characteristic matrix and the classification characteristic matrix by a matrix to obtain the corrected classification characteristic matrix.

9. The electric gate valve with monitoring system of claim 8, wherein the node-like topology-class center fusion unit is further configured to: performing node-like topology-like center fusion on the first diagonal matrix and the second diagonal matrix by using the following formula to obtain the fusion feature matrix; wherein, the formula is:

10. The electric gate valve with monitoring system of claim 9, wherein the monitoring result generation module comprises: the unfolding unit is used for unfolding the corrected classification characteristic matrix into a classification characteristic vector according to a row vector or a column vector; the full-connection coding unit is used for carrying out full-connection coding on the classification characteristic vectors by using a full-connection layer of the classifier so as to obtain coded classification characteristic vectors; and the classification result generating unit is used for inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.