Parametric fault diagnosis method for power electronic equipment and computer equipment

Abstract

The invention discloses a parametric fault diagnosis method for power electronic equipment and computer equipment, and the method comprises the steps: collecting a current signal of the power electronic equipment, and carrying out the normalization processing of the current signal through a normalization formula, and obtaining a normalized feature value; solving an arc cosine value of the characteristic value; constructing an Euler difference formula through an Euler formula, and substituting the arc cosine value corresponding to the characteristic value into the Euler difference formula to obtain an Euler difference characteristic value; the Euler differential feature value is converted into a grey-scale map, and the grey-scale map is used as an input feature image of a Vision Transform model; and classifying the input feature image by using a Vision Transform model to obtain an image category as a diagnosis result of different types of parametric faults of the power electronic equipment. According to the method, accurate diagnosis of the parametric fault of the power electronic equipment can be realized, and the problem that the parametric fault diagnosis of the power electronic equipment is inaccurate due to the fact that the characteristic change between different degrees of aging of the power electronic equipment is not obvious and degradation characteristics are difficult to effectively extract is solved.

Description

Technical Field

[0001] The invention belongs to the field of equipment fault diagnosis, and in particular relates to a method for diagnosing parameter faults of power electronic equipment and computer equipment. Background Art

[0002] Power electronic equipment is widely used in production and life, and its stability is of great significance to the safety of people's lives and property as well as industrial production. However, in actual operation, the working environment of power electronic equipment is very complex, which causes various devices in power electronic equipment to often experience huge operating pressure, resulting in various faults in power electronic devices. Timely monitoring of the status of power electronic devices and implementing pre-maintenance on the basis of planned maintenance can reduce the possibility of equipment failure and downtime, which has important practical significance.

[0003] In the actual working process, power electronic equipment may experience open circuit faults, short circuit faults, and aging failure of power devices due to the influence of environment and external stress. Among them, open circuit faults and short circuit faults will cause obvious changes in the voltage parameters of power electronic devices in the equipment. Especially when power electronic devices have aging faults, the parameters of the device will drift, resulting in parametric failures in the equipment. However, the fault characteristics of parametric failures are relatively small, which makes it difficult for general diagnostic methods to achieve accurate diagnosis of parametric failures of power electronic equipment.

[0004] In the existing electronic circuit fault diagnosis field, traditional time-frequency domain feature analysis methods such as Fourier transform, EMD (Empirical Mode Decomposition) decomposition, wavelet analysis, Hilbert-Huang transform, etc. only perform simple digital feature analysis and calculation on the output voltage or current analog quantity of the circuit. In recent years, with the development of artificial intelligence and machine vision technology, machine vision detection based on artificial intelligence algorithms has gradually gained attention in the field of fault diagnosis. However, most of the current machine vision fault diagnosis is still aimed at some appearance defects and abnormal equipment or component failures. It seems to be stretched for power electronic circuits with weak, complex features or unclear fault phenomena. The existing technical solutions based on neural networks usually extract features and then input the extracted features into a classification network to identify the fault mode. However, the classification network is usually simple and has a good diagnostic effect when facing open and short circuit faults of power electronic equipment. However, when the research object is the degradation of electronic devices, due to the small difference in features between different degradation degrees, the existing technical solutions are difficult to accurately judge the parametric faults of power electronic equipment. Summary of the invention

[0005] The purpose of the present invention is to provide a method for diagnosing parametric faults of power electronic equipment, a computer device, a computer-readable storage medium and a computer program product, which can realize accurate diagnosis of parametric faults of power electronic equipment, in order to solve the problem of inaccurate diagnosis of parametric faults of power electronic equipment caused by the unclear characteristic changes between different degrees of aging of power electronic equipment and the difficulty in effectively extracting degradation characteristics.

[0006] In order to achieve the above object, one aspect of the present invention provides a method for diagnosing parametric faults of power electronic equipment, comprising:

[0007] Step S1, collecting the current signal of the power electronic device, and normalizing the current signal by the following normalization formula to obtain the normalized characteristic value:

[0008]

[0009] Where X represents the set of current signals collected, max(X) represents the maximum value in the set, min(X) represents the minimum value in the set, and x i Represents a data point in the set, represents the normalized eigenvalue;

[0010] Step S2, calculate the arc cosine value of the eigenvalue: in, φ represents the arccosine value corresponding to the eigenvalue;

[0011] Step S3, constructing the Euler difference formula through the Euler formula: Substitute the arccosine value corresponding to the eigenvalue into the Euler difference formula to obtain the Euler difference eigenvalue;

[0012] Step S4, converting the Euler difference eigenvalue into a grayscale image as an input feature image of the Vision Transformer model;

[0013] Step S5, using the Vision Transformer model to classify the input feature image, and obtain image categories as diagnosis results of different types of parametric faults of power electronic equipment.

[0014] Preferably, step S5 comprises:

[0015] The input feature image is evenly cut into image blocks, flatten each image block into a one-dimensional vector, and the dimension of each vector is P 2 ×C, where H, W and C represent the length, width and height of the input feature image respectively, and P represents the size of the image block;

[0016] Add a position code to each image block to indicate its position in the input feature image, which is used to calculate the attention between image blocks;

[0017] Using the Transformer's self-attention mechanism, the attention between image blocks is calculated to obtain the encoding of each image block;

[0018] The obtained image block encoding is input into the fully connected layer of the multi-layer perceptron. After dimension conversion, the Softmax function is used to obtain the probability vector of the image belonging to each category. The maximum value is taken as the image category, which is used as the diagnosis result of different types of parametric faults of power electronic equipment.

[0019] Another aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above method.

[0020] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above method when executed by a processor.

[0021] Another aspect of the present invention provides a computer program product, comprising a computer program, which implements the steps of the above method when executed by a processor.

[0022] The power electronic equipment parametric fault diagnosis method, computer equipment, computer readable storage medium and computer program product according to the above aspects of the present invention can realize accurate diagnosis of power electronic equipment parametric faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings used in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work:

[0024] Figure 1 is a flow chart of a method for diagnosing parametric faults of power electronic equipment according to an embodiment of the present invention;

[0025] Figure 2 is a schematic diagram of a Vision Transformer model according to an embodiment of the present invention;

[0026] Figure 3 It is a structural diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The embodiment of the present invention aims at the problems of inaccurate parametric fault diagnosis and low recognition accuracy of power electronic equipment caused by the fact that the characteristic changes between different degrees of aging of power electronic equipment are not obvious in the prior art solution, and conventional methods are difficult to effectively extract degradation characteristics, thereby proposing a parametric fault diagnosis method for power electronic equipment based on AED (Arccos Euler Difference) and Vision Transformer. In this embodiment, the power electronic equipment is, for example, a DC-DC Boost circuit (boost chopper circuit).

[0029] Figure 1 FIG. 1 is a flow chart of a method for diagnosing parametric faults of power electronic equipment according to an embodiment of the present invention. Figure 1 As shown, the method for diagnosing parametric faults of power electronic equipment according to the embodiment of the present invention includes steps S1 to S5.

[0030] In step S1, the current signal of the power electronic circuit is collected, and the current signal is first normalized by a normalization formula to obtain a normalized eigenvalue:

[0031]

[0032] Where X represents the collection of the original current signals collected, which is the overall data, x i is the target data point, max(X) represents the maximum value in the set, and min(X) represents the minimum value in the set. Represents the normalized eigenvalue.

[0033] In step S2, the arc cosine value of the eigenvalue is calculated: in, φ represents the arccosine value corresponding to the characteristic value. In this step, the arccosine value in the current signal is extracted to realize the utilization of the time series feature.

[0034] In step S3, the Euler difference formula is constructed by the Euler formula: The arc cosine value corresponding to the eigenvalue is substituted into the Euler difference formula to output the Euler difference eigenvalue. In this step, the arc cosine value is substituted into the Euler difference formula, and the spatial features in the signal are extracted by the Euler difference formula, so as to highlight the fault features in the characteristic signal and suppress the noise signal.

[0035] In step S4, the Euler difference eigenvalue is directly converted into a grayscale image as the input feature image of the Vision Transformer. In this step, the grayscale image conversion is to directly draw the data waveform after the original data is extracted by AED, with the waveform part in black and the remaining part in white. The feature image generated by AED not only maintains the small size of the grayscale image, but also mines the fault features in the feature signal, which is conducive to the extraction of fault features by the diagnostic model and improves the diagnostic accuracy of the model.

[0036] In step S5, the input feature image is classified using the Vision Transformer model to obtain image categories as diagnosis results of different types of parametric faults of power electronic equipment.

[0037] Vision Transformer (ViT) is a diagnostic model that applies Transformer to image classification. The feature highlighting capability of AED is combined with the feature mining capability of Vision Transformer to achieve high-precision fault diagnosis of parametric faults in power electronic equipment.

[0038] Let the length, width and height of the input image be H, W and C respectively, and P represents the size of the image block. The ViT model architecture is as follows Figure 2 As shown, the above step S5 specifically includes the following steps.

[0039] Step 1: Cut the input feature image into pieces. First, cut the input feature image into pieces evenly. image blocks, the image blocks a to i in the figure. Then, each two-dimensional image block is flattened into a one-dimensional vector, and the dimension of each vector is P 2 ×C.

[0040] Step 2: Position encoding. Encode each image block, map each flattened image block to a D-dimensional vector, and add a position code to each image block to indicate its original position in the image, which is used to calculate the attention between image blocks.

[0041] Step 3: Self-attention encoding. Using the self-attention mechanism in Transformer, we use multi-head self-attention (Multi-Head Attention), multi-layer perceptron (MLP), layer normalization (Norm) and other operations to calculate the attention between image blocks and obtain the encoding of each image block.

[0042] Step 4: Calculate the image category. The image block encoding obtained by the Transformer encoder is input into the fully connected layer of the MLP. After dimension conversion, the Softmax function is used to obtain the probability vector of the image belonging to each category. The maximum value is taken to obtain the image category as the diagnosis result of different types of parametric faults of power electronic equipment.

[0043] The Transformer multi-head attention mechanism is a sequence-based network architecture. Given a query element query and a set of key elements, the multi-head attention mechanism weighs the attention weights to adaptively aggregate the compatibility of key and query key pairs. Let q represent the query element, and its features are recorded as k∈Ω k Represents the key element, whose characteristics are recorded as Where C is the feature dimension, Ω q and Ω k Represent the collection of query elements and key elements respectively. The formula of the multi-head attention mechanism is: Where: M is the number of heads of multi-head attention, All are learnable parameters, the default is C V =C / M. Attention weight Normalized to Among them U m and It is also a learnable parameter.

[0044] In summary, the method of the embodiment of the present invention constructs a parametric fault diagnosis method for power electronic equipment through AED, grayscale image conversion and VisionTransformer. The output current signal of the power electronic circuit is collected, the arc cosine value corresponding to each sampling point is obtained after normalization, and the eigenvalue is obtained through the Euler difference formula. After the eigenvalue conversion by AED, the fault characteristics in the current signal are highlighted and the noise signal is suppressed. Finally, the eigenvalue is converted into a grayscale image as the input feature image of Vision Transformer, so as to realize accurate diagnosis of parametric faults of power electronic equipment.

[0045] Compared with other fault diagnosis methods for power electronic equipment, the method of the embodiment of the present invention has the following advantages:

[0046] 1) The AED is used to extract the time series features of the original current signal, highlight the fault features in the fault signal, and suppress the noise;

[0047] 2) The characteristic signal after AED is directly converted into a grayscale image as the input feature image of the diagnostic model. The grayscale image generated by AED avoids the problem of slow model training speed caused by the huge size of the image generated by other image conversion methods, and retains the fault characteristics in the signal, so that the model has better diagnostic accuracy;

[0048] 3) The feature extraction method AED is combined with Vision Transformer to maximize the mining of fault features in feature images that can represent different types of parametric faults.

[0049] An embodiment of the present invention further provides a computer device, which may be a server, and its internal structure diagram may be as shown in FIG. Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store operating parameter data of each framework. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps of the method of the embodiment of the present invention are implemented.

[0050] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0051] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method of the embodiment of the present invention are implemented.

[0052] An embodiment of the present invention further provides a computer program product, including a computer program, which implements the steps of the method of the embodiment of the present invention when executed by a processor.

[0053] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for diagnosing parametric faults of power electronic equipment, characterized in that: include: Step S1, collecting the current signal of the power electronic device, and normalizing the current signal by the following normalization formula to obtain the normalized characteristic value: Where X represents the set of current signals collected, max(X) represents the maximum value in the set, min(X) represents the minimum value in the set, and x i Represents a data point in the set, represents the normalized eigenvalue; Step S2, calculate the arc cosine value of the eigenvalue: in, φ represents the arccosine value corresponding to the eigenvalue; Step S3, constructing the Euler difference formula through the Euler formula: Substitute the arccosine value corresponding to the eigenvalue into the Euler difference formula to obtain the Euler difference eigenvalue; Step S4, converting the Euler difference eigenvalue into a grayscale image as an input feature image of the Vision Transformer model; Step S5, using the Vision Transformer model to classify the input feature image, and obtain image categories as diagnosis results of different types of parametric faults of power electronic equipment.

2. The method according to claim 1, characterized in that Step S5 includes: Cut the input feature image evenly into image blocks, flatten each image block into a one-dimensional vector, and the dimension of each vector is P 2 ×C, where H, W and C represent the length, width and height of the input feature image respectively, and P represents the size of the image block; Add a position code to each image block to indicate its position in the input feature image, which is used to calculate the attention between image blocks; Using the Transformer's self-attention mechanism, the attention between image blocks is calculated to obtain the encoding of each image block; The obtained image block encoding is input into the fully connected layer of the multi-layer perceptron. After dimension conversion, the Softmax function is used to obtain the probability vector of the image belonging to each category. The maximum value is taken as the image category, which is used as the diagnosis result of different types of parametric faults of power electronic equipment.

3. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to claim 1 or 2.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claim 1 or 2 are implemented.

5. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to claim 1 or 2 are implemented.

Images