Power equipment fault detection method and system based on deep learning network

Abstract

The invention provides a power equipment fault detection method and system based on a deep learning network, and the method comprises the steps: obtaining state parameter data of power equipment in an observation time period, forming a first multi-dimensional time sequence according to the state parameter data, inputting the first multi-dimensional time sequence to a deep learning model, obtaining a fault probability prediction value corresponding to each time node; obtaining a fault coefficient according to the fault probability predicted values of all the time nodes, and judging whether the power equipment has a fault according to the fault coefficient; and if the power equipment has a fault, inputting the fault probability prediction value corresponding to each time node into the fault classification model to obtain a fault category set corresponding to the observation time period, and obtaining the fault category of the current time node according to the fault category set. According to the method and the device, the condition of inaccurate diagnosis result caused by data exception due to fault judgment through single-dimension data can be avoided, and the fault diagnosis accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of fault detection technology, and in particular to a method and system for detecting faults of electric power equipment based on a deep learning network. Background Art

[0002] In the power system, the stable operation of power equipment is crucial to ensure the safety of the system. Fault detection of power equipment is carried out to promptly know whether the power equipment has failed. When the power equipment fails, it needs to be repaired to prevent the failed power equipment from continuing to operate and eliminate safety hazards. However, in actual applications, the complexity of the equipment and the diversity of the operating environment make it difficult to detect equipment failures.

[0003] In order to improve the accuracy of fault detection, the invention application with publication number CN115979456A in the prior art discloses a fault detection method, device and equipment for power equipment, which collects numerical values ​​of multiple target dimensions of power equipment and detects the fault level corresponding to the power equipment, that is, uses machine learning technology to detect the fault level corresponding to the power equipment, which can improve the accuracy of the fault detection results of the power equipment.

[0004] In practical applications, overheating of power equipment may be caused by faults, but the fault condition of power equipment is not only related to temperature. Feature data of other dimensions can also reflect related faults of power equipment. For example, for motor fault diagnosis, vibration sensors can detect machine vibrations, and machine vibration data reflects mechanical imbalance faults. Current and voltage sensors can detect the current and voltage of power equipment, and current and voltage data can reflect electrical faults. Multi-dimensional data generally have a corresponding threshold set. When the data exceeds the threshold, it can only be judged that the data is abnormal, and it cannot be directly judged whether there is a fault in the power equipment. The method of fault judgment through thresholds is relatively simple, and the accuracy of fault diagnosis is not high. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for detecting faults in power equipment based on a deep learning network, aiming to solve the problem that traditional fault detection technology can only directly conclude that the data will be abnormal by setting thresholds, but cannot directly determine whether there is a fault in the power equipment, the method of fault judgment is relatively single, and the fault diagnosis accuracy is not high.

[0006] In a first aspect, the present invention provides a method for detecting faults of electric power equipment based on a deep learning network, the method comprising: Build a deep learning model consisting of a memory module and a convolutional neural network, and build a fault classification model based on the convolutional neural network; An observation period is selected to obtain state parameter data of the power equipment during the observation period, and a first multidimensional time series is formed according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and the first multidimensional time series is input into the deep learning model to obtain a fault probability prediction value corresponding to each time node; Obtaining a fault coefficient according to the predicted values ​​of the fault probabilities at all time nodes, and judging whether the power equipment has a fault according to the fault coefficient; If there is a fault in the power equipment, the fault probability prediction value corresponding to each time node is input into the fault classification model to obtain a set of fault categories corresponding to the observation period, and the fault category of the current time node is obtained according to the set of fault categories.

[0007] Furthermore, the steps of constructing a deep learning model consisting of a memory module and a convolutional neural network, and constructing a fault classification model based on the convolutional neural network include: At any time point t, historical data of each state parameter of the power equipment from time point tn to t are collected to form a second multidimensional time series; The second multi-dimensional time series is input into the memory module, subjected to time dimension feature extraction and output into the convolutional neural network, and local feature extraction and high-level feature extraction are sequentially performed to obtain a fault probability prediction value; Adjust the weights of each layer in the memory module and the convolutional neural network according to the predicted value of the fault probability.

[0008] Furthermore, the step of inputting the second multidimensional time series into the memory module, extracting the features of the time dimension and outputting them into the convolutional neural network, and sequentially performing local feature extraction and high-level feature extraction to obtain the fault probability prediction value includes: In the memory module, the feature extraction process of the time dimension is as follows: The forget gate is calculated by the following formula: ; The candidate cell states at the input gate and the current time point t are calculated by the following formula: ; ; The cell state at the current time point t is updated using the forget gate, input gate, and candidate cell state through the following formula: ; The output gate is calculated and the hidden state at the current time t is updated by the following formula: ; ; In the formula, For the Gate of Forgetfulness, is the input gate, is the candidate cell state at time point t, is the cell state at time point t, is the output gate, is the hidden state at time point t, is the sigmoid function, is the hidden state at time point t-1, is the real-time data of the state parameters at time point t, , , , , , , , are all learnable parameters in the memory module, tanh is the activation function, is the cell state at time point t-1.

[0009] Furthermore, the step of inputting the second multidimensional time series into the memory module, extracting the features of the time dimension and outputting them into the convolutional neural network, and sequentially performing local feature extraction and high-level feature extraction to obtain the fault probability prediction value also includes: Substitute the hidden state at time point t into a convolutional neural network, where the convolutional neural network is set to have two layers; In the first layer of the convolutional neural network, the hidden state at time point t is convolved and nonlinearly changed using the following formula to obtain the output value: ; ; In the formula, , is the learnable parameter of the first layer of convolutional neural network, is the first convolution value, A is the output value, and ReLu is the activation function; In the second layer of the convolutional neural network, the following formula is used to perform a convolution operation on the output value and convert the convolution value into a fault probability prediction value based on the softmax function: ; ; In the formula, , is the learnable parameter of the second layer of convolutional neural network, is the second convolution value, and p is the predicted value of the failure probability.

[0010] Furthermore, the step of obtaining a fault coefficient according to the fault probability prediction values ​​of all time nodes, and judging whether the power equipment has a fault according to the fault coefficient includes: The failure coefficient is obtained according to the following formula: ; In the formula, is the predicted value of the failure probability at the tth time point, T represents the observation period, and Y represents the failure coefficient during the observation period.

[0011] Determine whether the failure coefficient during the observation period is greater than a preset failure coefficient threshold; If the fault coefficient during the observation period is greater than a preset fault coefficient threshold, it is determined that a fault exists in the power equipment.

[0012] Furthermore, if the power equipment has a fault, the fault probability prediction value corresponding to each time node is input into the fault classification model to obtain a set of fault categories corresponding to the observation period, and the step of obtaining the fault category of the current time node according to the set of fault categories includes: The fault probability prediction value at each time point based on the convolutional neural network is mapped to the fault category parameter u through a fully connected layer: ; In the formula, and are all parameters of the fully connected layer, and p is the predicted value of the fault probability; The fault category parameter u is converted into the probability of different fault categories through the softmax function: ; In the formula, represents the fault probability of the jth fault category at the time point corresponding to the fault category parameter u, c is the total number of fault categories, and exp represents the natural exponential function; Screening out the maximum fault probability from all fault probabilities at the same time point, and taking the fault category corresponding to the maximum fault probability as the final fault category at the time point; Output the final failure probability at all time points during the observation period.

[0013] Furthermore, the method further comprises: In the training process of memory module and convolutional neural network, the objective function is defined as:

[0014] In the formula, is the total number of samples, is the actual status of the power equipment, is the predicted value of the failure probability obtained by the deep learning model, Indicates that the actual state of the power equipment at time point t is normal, Indicates that the actual state of the power equipment at time point t is abnormal.

[0015] In a second aspect, the present invention provides a power equipment fault detection system based on a deep learning network, the system comprising: A model building module is used to build a deep learning model consisting of a memory module and a convolutional neural network, and to build a fault classification model based on the convolutional neural network; A probability prediction module is used to select an observation period to obtain state parameter data of the power equipment within the observation period, and form a first multidimensional time series according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and input the first multidimensional time series into the deep learning model to obtain a fault probability prediction value corresponding to each time node; A fault detection module, used to obtain a fault coefficient according to the fault probability prediction values ​​of all time nodes, and determine whether the power equipment has a fault according to the fault coefficient; The fault category output module is used to input the fault probability prediction value corresponding to each time node into the fault classification model if there is a fault in the power equipment, obtain the fault category set corresponding to the observation period, and obtain the fault category of the current time node according to the fault category set.

[0016] In a third aspect, the present invention provides a storage medium, which stores one or more programs, which, when executed by a processor, implement the above-mentioned power equipment fault detection method based on deep learning network.

[0017] In a fourth aspect, the present invention provides an electronic device, the electronic device comprising a memory and a processor, wherein: The memory is used to store computer programs; When the processor is used to execute the computer program stored in the memory, the above-mentioned power equipment fault detection method based on deep learning network is implemented.

[0018] In summary, according to the above-mentioned power equipment fault detection method based on deep learning network, the deep learning model is trained through the historical data of various state parameters of the power equipment, and the fault probability prediction value of the fault and the fault coefficient of the power equipment are accurately calculated according to the deep learning model. If the fault coefficient exceeds the threshold, further fault classification detection is performed, and the fault coefficient of the power equipment can be calculated through multi-dimensional state parameters. The fault coefficient is used as a judgment standard for the current operating status of the power equipment. If the operating status is not good, further fault classification detection is performed, which improves the efficiency of the fault diagnosis process. In addition, the fault classification model is combined with the fault prediction probability obtained based on the multi-dimensional state parameter analysis to jointly perform comprehensive classification diagnosis on the fault, avoiding the situation where fault judgment is made through single-dimensional data and the diagnosis result is inaccurate due to data abnormality, thereby improving the accuracy of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flowchart of a method for detecting faults in electric power equipment based on a deep learning network proposed in one embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a power equipment fault detection system based on a deep learning network proposed in one embodiment of the present invention.

[0020] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0022] like Figure 1 As shown, an embodiment of the present invention provides a method for detecting faults of electric power equipment based on a deep learning network, the method comprising steps S101 to S104, wherein: Step S101: constructing a deep learning model consisting of a memory module and a convolutional neural network, and constructing a fault classification model based on the convolutional neural network; It should be noted that, in some embodiments, during the training of the memory module and the convolutional neural network, the objective function is defined as: ; In the formula, N is the total number of samples, is the actual status of the power equipment, is the fault probability prediction value obtained by the deep learning model. The optimization goal of the deep learning model is to minimize the gap between the fault probability prediction value and the actual state of the power equipment. After optimizing the deep learning model using the above objective function, the fault probability prediction value is calculated using actual data, and the mean value of the fault probability prediction value is obtained to obtain the fault coefficient of the power equipment.

[0023] In addition, in some other embodiments of the present invention, during the training process of the memory module and the convolutional neural network, the objective function is defined as:

[0024] In the formula, is the total number of samples, is the actual status of the power equipment, is the predicted value of the failure probability obtained by the deep learning model, Indicates that the actual state of the power equipment at time point t is normal, Indicates that the actual state of the power equipment at time point t is abnormal. During the training process, if the actual state of the power equipment is normal, the goal is to minimize the residual, that is, to minimize the gap between the fault probability prediction value and the actual state of the power equipment. If the actual state of the power equipment is abnormal, the goal is to maximize the residual, that is, to maximize the gap between the fault probability prediction value and the actual state of the power equipment. After optimizing the deep learning model using the above objective function, the fault probability prediction value of the abnormal actual state of the power equipment can be effectively monitored, so that the data of the fault probability prediction value under the abnormal state of the power equipment is increased. In this case, the fault probability prediction value can be calculated more accurately, and the mean of the fault probability prediction value can be obtained to obtain the failure coefficient of the power equipment.

[0025] Step S102: selecting an observation period to obtain state parameter data of the power equipment within the observation period, and forming a first multidimensional time series according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and inputting the first multidimensional time series into the deep learning model to obtain a fault probability prediction value corresponding to each time node; It should be pointed out that the state parameter data includes but is not limited to temperature data, vibration amplitude, vibration frequency, current and voltage data of the power equipment.

[0026] It should be noted that the fault prediction situations in different observation periods may be different. For example, in actual application, the equipment was replaced three months ago. The historical data of the past three months has reference value, but the previous data has no reference value for the fault prediction results. The accuracy of the obtained probability prediction value is far lower than the value taken in the past three months. Therefore, in actual application, according to the actual situation, only the historical data of the past three months is collected for training the deep learning model.

[0027] In addition, for the structure of different power equipment, it is also necessary to select different continuous time periods to calculate the failure coefficient of the corresponding power equipment. For example, for the vibration of the bearing seat, a continuous period of 2h can be selected as the observation period, and the probability prediction value is calculated at each time interval during the observation period. The average of the probability prediction values ​​is used to obtain the failure coefficient of the power equipment. A longer time is selected as the observation period, and the change of the state parameter will not have a big impact on the whole, but will only have an impact under long-term abnormal operation. It can be reflected in time when the failure coefficient is abnormal during the observation period on site.

[0028] For example, if the corresponding current and voltage are abnormal, a continuous period of 10s can be selected as the observation period, and the probability prediction value can be calculated at each time interval (for example, 1s) within the observation period, and the average of the probability prediction values ​​can be obtained to obtain the failure coefficient of the power equipment. A shorter time is selected as the observation period. The change of this state parameter will have a greater impact on the whole. Short-term abnormalities will affect the operation of the power equipment. The observation period is short, and the failure coefficient of the power equipment in a short period of time can be calculated and reflected in a timely and effective manner.

[0029] Step S103: obtaining a fault coefficient according to the fault probability prediction values ​​of all time nodes, and judging whether the power equipment has a fault according to the fault coefficient; Specifically, a continuous period T including the current time point is selected as the observation period, and the probability prediction value is calculated at each interval within the observation period T, and the average value of the probability prediction value is obtained to obtain the failure coefficient of the power equipment: ; In the formula, is the predicted value of the failure probability at the tth time point, T represents the observation period, and Y represents the failure coefficient during the observation period.

[0030] Then, determining whether the fault coefficient in the observation period is greater than a preset fault coefficient threshold; If the fault coefficient in the observation period is greater than a preset fault coefficient threshold, it is determined that the power equipment has a fault; If the fault coefficient during the observation period is less than or equal to the preset fault coefficient threshold, it is determined that there is no fault in the power equipment, and there is no need to enter the fault classification.

[0031] Step S104: If the power equipment has a fault, the fault probability prediction value corresponding to each time node is input into the fault classification model to obtain a set of fault categories corresponding to the observation period, and the fault category of the current time node is obtained according to the set of fault categories.

[0032] It should be pointed out that in the fault classification model, the fault probability prediction value at each time point based on the convolutional neural network is mapped to the fault category parameter u through the fully connected layer: ; In the formula, and are all parameters of the fully connected layer, and p is the predicted value of the fault probability; The fault category parameter u is converted into the probability of different fault categories through the softmax function: ; In the formula, represents the fault probability of the jth fault category at the time point corresponding to the fault category parameter u, c is the total number of fault categories, and exp represents the natural exponential function; The maximum fault probability is selected from all fault probabilities at the same time point, and the fault category corresponding to the maximum fault probability is used as the final fault category at the time point; and the final fault probability of all time points in the observation period is finally output. In other words, through the method proposed in this step, the relationship between the fault category parameter data and the fault category can be obtained at a certain time point. The first time point can obtain a fault category A, the second time point can obtain a fault category B... and so on, thereby obtaining a one-to-one correspondence between the state parameter data and the fault category in continuous time, thereby achieving the purpose of fault classification.

[0033] In addition, in some embodiments, in the process of building a deep learning model, first, for any time point t, the historical data of each state parameter of the power equipment from time point tn to t is collected to form a second multidimensional time series; then the second multidimensional time series is input into the memory module, and after the feature extraction of the time dimension, it is output into the convolutional neural network, and local feature extraction and high-level feature extraction are performed in turn to obtain the fault probability prediction value; finally, the weights of each layer in the memory module and the convolutional neural network are adjusted according to the fault probability prediction value. Specifically, in the memory module, the feature extraction process of the time dimension is as follows: The forget gate is calculated by the following formula: ; The candidate cell states at the input gate and the current time point t are calculated by the following formula: ; ; The cell state at the current time point t is updated using the forget gate, input gate, and candidate cell state through the following formula: ; The output gate is calculated and the hidden state at the current time t is updated by the following formula: ; ; In the formula, For the Gate of Forgetfulness, is the input gate, is the candidate cell state at time point t, is the cell state at time point t, is the output gate, is the hidden state at time point t, is the sigmoid function, is the hidden state at time point t-1, is the real-time data of the state parameters at time point t, , , , , , , , are all learnable parameters in the memory module, tanh is the activation function, is the cell state at time point t-1.

[0034] In addition, in the construction of the fault classification model, the hidden state at time point t needs to be substituted into the convolutional neural network, and the convolutional neural network is set to two layers; In the first layer of the convolutional neural network, the hidden state at time point t is convolved and nonlinearly changed using the following formula to obtain the output value: ; ; In the formula, , is the learnable parameter of the first layer of convolutional neural network, is the first convolution value, A is the output value, and ReLu is the activation function; In the second layer of the convolutional neural network, the following formula is used to perform a convolution operation on the output value and convert the convolution value into a fault probability prediction value based on the softmax function: ; ; In the formula, , is the learnable parameter of the second layer of convolutional neural network, is the second convolution value, and p is the predicted value of the failure probability.

[0035] By constructing two new algorithms, namely a deep learning model and a fault classification model, both models use the historical data of power equipment as the original training set. After collecting data from a specific observation period, the two models work closely together to efficiently and accurately analyze whether there is a fault and the type of fault that exists.

[0036] In summary, according to the above-mentioned power equipment fault detection method based on deep learning network, the deep learning model is trained through the historical data of various state parameters of the power equipment, and the fault probability prediction value of the fault and the fault coefficient of the power equipment are accurately calculated according to the deep learning model. If the fault coefficient exceeds the threshold, further fault classification detection is performed, and the fault coefficient of the power equipment can be calculated through multi-dimensional state parameters. The fault coefficient is used as a judgment standard for the current operating status of the power equipment. If the operating status is not good, further fault classification detection is performed, which improves the efficiency of the fault diagnosis process. In addition, the fault classification model is combined with the fault prediction probability obtained based on the multi-dimensional state parameter analysis to jointly perform comprehensive classification diagnosis on the fault, avoiding the situation where fault judgment is made through single-dimensional data and the diagnosis result is inaccurate due to data abnormality, thereby improving the accuracy of fault diagnosis.

[0037] like Figure 2 As shown, an embodiment of the present invention further provides a power equipment fault detection system based on a deep learning network, the system comprising: A model building module 10 is used to build a deep learning model consisting of a memory module and a convolutional neural network, and to build a fault classification model based on the convolutional neural network; The probability prediction module 20 is used to select an observation period to obtain state parameter data of the power equipment in the observation period, and form a first multidimensional time series according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and input the first multidimensional time series into the deep learning model to obtain a fault probability prediction value corresponding to each time node; A fault detection module 30 is used to obtain a fault coefficient according to the fault probability prediction values ​​of all time nodes, and to determine whether the power equipment has a fault according to the fault coefficient; The fault category output module 40 is used to input the fault probability prediction value corresponding to each time node into the fault classification model if there is a fault in the power equipment, obtain a fault category set corresponding to the observation period, and obtain the fault category of the current time node according to the fault category set.

[0038] On the other hand, the present invention further proposes a storage medium on which one or more programs are stored, which, when executed by a processor, implement the above-mentioned power equipment fault detection method based on deep learning network.

[0039] On the other hand, the present invention also proposes an electronic device, including a memory and a processor, wherein the memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the above-mentioned power equipment fault detection method based on deep learning network.

[0040] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain storage, communication, propagation or transmission of a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0041] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0042] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0043] Although the embodiments of the present invention are described in detail above, it is obvious to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as described in the claims. Moreover, the present invention described herein may have other embodiments and may be implemented or realized in a variety of ways.

Claims

1. A method for detecting faults in power equipment based on a deep learning network, characterized in that: The method comprises: Build a deep learning model consisting of a memory module and a convolutional neural network, and build a fault classification model based on the convolutional neural network; An observation period is selected to obtain state parameter data of the power equipment during the observation period, and a first multidimensional time series is formed according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and the first multidimensional time series is input into the deep learning model to obtain a fault probability prediction value corresponding to each time node; Obtaining a fault coefficient according to the predicted values ​​of the fault probabilities at all time nodes, and judging whether the power equipment has a fault according to the fault coefficient; If there is a fault in the power equipment, the fault probability prediction value corresponding to each time node is input into the fault classification model to obtain a set of fault categories corresponding to the observation period, and the fault category of the current time node is obtained according to the set of fault categories.

2. The method for detecting faults of electric power equipment based on deep learning network according to claim 1, characterized in that: The steps of constructing a deep learning model consisting of a memory module and a convolutional neural network, and constructing a fault classification model based on the convolutional neural network include: At any time point t, historical data of each state parameter of the power equipment from time point tn to t are collected to form a second multidimensional time series; The second multi-dimensional time series is input into the memory module, subjected to time dimension feature extraction and output into the convolutional neural network, and local feature extraction and high-level feature extraction are sequentially performed to obtain a fault probability prediction value; Adjust the weights of each layer in the memory module and the convolutional neural network according to the predicted value of the fault probability.

3. The method for detecting faults of electric power equipment based on deep learning network according to claim 2, characterized in that: The step of inputting the second multi-dimensional time series into the memory module, extracting the time dimension features and outputting them into the convolutional neural network, and sequentially performing local feature extraction and high-level feature extraction to obtain the fault probability prediction value comprises: In the memory module, the feature extraction process of the time dimension is as follows: The forget gate is calculated by the following formula: ； The candidate cell states at the input gate and the current time point t are calculated by the following formula: ； ； The cell state at the current time point t is updated using the forget gate, input gate, and candidate cell state through the following formula: ； The output gate is calculated and the hidden state at the current time t is updated by the following formula: ； ； In the formula, For the Gate of Forgetfulness, is the input gate, is the candidate cell state at time point t, is the cell state at time point t, is the output gate, is the hidden state at time point t, is the sigmoid function, is the hidden state at time point t-1, is the real-time data of the state parameters at time point t, , , , , , , , are all learnable parameters in the memory module, tanh is the activation function, is the cell state at time point t-1.

4. The method for detecting faults of electric power equipment based on deep learning network according to claim 3, characterized in that: The step of inputting the second multi-dimensional time series into the memory module, extracting the time dimension features and outputting them into the convolutional neural network, and sequentially performing local feature extraction and high-level feature extraction to obtain the fault probability prediction value also includes: Substitute the hidden state at time point t into a convolutional neural network, where the convolutional neural network is set to have two layers; In the first layer of the convolutional neural network, the hidden state at time point t is convolved and nonlinearly changed using the following formula to obtain the output value: ； ； In the formula, , is the learnable parameter of the first layer of convolutional neural network, is the first convolution value, A is the output value, and ReLu is the activation function; In the second layer of the convolutional neural network, the following formula is used to perform a convolution operation on the output value and convert the convolution value into a fault probability prediction value based on the softmax function: ； ； In the formula, , is the learnable parameter of the second layer of convolutional neural network, is the second convolution value, and p is the predicted value of the failure probability.

5. The method for detecting faults of electric power equipment based on deep learning network according to claim 1, characterized in that: The step of obtaining a fault coefficient according to the fault probability prediction values ​​of all time nodes, and judging whether the power equipment has a fault according to the fault coefficient comprises: The failure coefficient is obtained according to the following formula: ； In the formula, is the predicted value of the failure probability at the tth time point, T represents the observation period, and Y represents the failure coefficient during the observation period; Determine whether the failure coefficient during the observation period is greater than a preset failure coefficient threshold; If the fault coefficient during the observation period is greater than a preset fault coefficient threshold, it is determined that a fault exists in the power equipment.

6. The method for detecting faults of electric power equipment based on deep learning network according to claim 1, characterized in that: If the power equipment has a fault, the fault probability prediction value corresponding to each time node is input into the fault classification model to obtain a set of fault categories corresponding to the observation period, and the step of obtaining the fault category of the current time node according to the set of fault categories includes: The fault probability prediction value at each time point based on the convolutional neural network is mapped to the fault category parameter u through a fully connected layer: ； In the formula, and are all parameters of the fully connected layer, and p is the predicted value of the fault probability; The fault category parameter u is converted into the probability of different fault categories through the softmax function: ； In the formula, represents the fault probability of the jth fault category at the time point corresponding to the fault category parameter u, c is the total number of fault categories, and exp represents the natural exponential function; Screening out the maximum fault probability from all fault probabilities at the same time point, and taking the fault category corresponding to the maximum fault probability as the final fault category at the time point; Output the final failure probability at all time points during the observation period.

7. The method for detecting faults of electric power equipment based on deep learning network according to claim 1, characterized in that: The method further comprises: In the training process of memory module and convolutional neural network, the objective function is defined as: ； In the formula, is the total number of samples, is the actual status of the power equipment, is the predicted value of the failure probability obtained by the deep learning model, Indicates that the actual state of the power equipment at time point t is normal, Indicates that the actual state of the power equipment at time point t is abnormal.

8. A power equipment fault detection system based on deep learning network, characterized in that: The system comprises: A model building module is used to build a deep learning model consisting of a memory module and a convolutional neural network, and to build a fault classification model based on the convolutional neural network; A probability prediction module is used to select an observation period to obtain state parameter data of the power equipment within the observation period, and form a first multidimensional time series according to the state parameter data, wherein the first multidimensional time series includes state parameter data of a current time node and at least one time node in the past, and input the first multidimensional time series into the deep learning model to obtain a fault probability prediction value corresponding to each time node; A fault detection module, used to obtain a fault coefficient according to the fault probability prediction values ​​of all time nodes, and determine whether the power equipment has a fault according to the fault coefficient; The fault category output module is used to input the fault probability prediction value corresponding to each time node into the fault classification model if there is a fault in the power equipment, obtain the fault category set corresponding to the observation period, and obtain the fault category of the current time node according to the fault category set.

9. A storage medium, characterized in that: The storage medium stores one or more programs, which, when executed by a processor, implement the power equipment fault detection method based on a deep learning network as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein: The memory is used to store computer programs; When the processor is used to execute a computer program stored in the memory, it implements the power equipment fault detection method based on a deep learning network as described in any one of claims 1 to 7.

Images