Fault processing method and device of switch cabinet and electronic equipment

Abstract

The invention discloses a fault processing method and device for a switch cabinet and electronic equipment, and relates to the field of electric power. The method comprises the following steps: acquiring equipment information of the switch cabinet; the equipment information of the switch cabinet is input into a fault detection model, fault detection is carried out on the switch cabinet based on the equipment information through the fault detection model, the state of the switch cabinet is obtained, and the fault detection model is a model obtained by adjusting parameters of a convolutional neural network based on a sparrow search algorithm. The states of the switch cabinet at least comprise a normal state, an insulation degradation state, a partial discharge state and a temperature rise standard exceeding state; generating a fault work order according to the state of the switch cabinet; and performing fault processing on the switch cabinet based on the fault work order. According to the invention, the technical problem of low fault processing efficiency of the switch cabinet is solved.

Description

Technical Field

[0001] The present invention relates to the field of electric power, and in particular to a fault handling method, device and electronic equipment for a switch cabinet. Background Art

[0002] With the increasingly stringent requirements of modern power systems for safe and stable operation, the health status of switchgear, as a key electrical equipment in the power grid, is directly related to the safety of the power grid and the reliability of power supply. Existing switchgear fault detection and processing solutions mainly rely on regular manual inspections and experience-based fault diagnosis. Inspection personnel regularly inspect switchgear according to the plan, record the operating status of the equipment and possible problems, and push alarm information to all operation and maintenance personnel through the data platform.

[0003] However, the above method relies on manual operation, has the problem of low fault detection efficiency, and the distribution accuracy of fault information is low. It is impossible to reasonably dispatch operation and maintenance resources according to the specific situation and urgency of the fault, resulting in low fault handling efficiency of the switch cabinet.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a switch cabinet fault handling method, device and electronic equipment to at least solve the technical problem of low efficiency in switch cabinet fault handling.

[0006] According to one aspect of an embodiment of the present invention, a fault handling method for a switch cabinet is provided, comprising: obtaining equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulation material degradation gas information, equipment operating life and location information; inputting the equipment information of the switch cabinet into a fault detection model, performing fault detection on the switch cabinet based on the equipment information through the fault detection model, and obtaining a state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting parameters of a convolutional neural network based on a sparrow search algorithm, and the state of the switch cabinet includes at least: a normal state, an insulation degradation state, a partial discharge state and a temperature rise exceeding standard state; generating a fault work order according to the state of the switch cabinet; and performing fault handling on the switch cabinet based on the fault work order.

[0007] In one implementation of the present application, fault processing of a switch cabinet based on a fault work order includes: classifying the fault work order to obtain a classification result of the fault work order; and performing fault processing on the switch cabinet based on the classification result.

[0008] In one implementation of the present application, fault work orders are classified to obtain classification results of the fault work orders, including: extracting features of the fault work orders to obtain a word vector matrix; inputting the word vector matrix into a classification model, and classifying the fault work orders based on the word vector matrix through the classification model to obtain classification results of the fault work orders.

[0009] In one implementation of the present application, feature extraction is performed on a fault work order to obtain a word vector matrix, including: performing word segmentation processing on the fault work order to obtain a vocabulary list of the fault work order; inputting the vocabulary list into a skip-gram model, and performing word vectorization processing on the vocabulary list through the skip-gram model to obtain a word vector matrix.

[0010] In one implementation of the present application, word segmentation is performed on a fault work order to obtain a vocabulary list of the fault work order, including: performing a data cleaning operation on the fault work order to obtain a fault work order after the data cleaning operation; performing word segmentation on the fault work order after the data cleaning operation to obtain data after word segmentation; performing deduplication processing on the data after word segmentation to obtain a vocabulary list of the fault work order.

[0011] In one embodiment of the present application, the hyperparameters in the fault detection model are determined based on a sparrow search algorithm, and the internal parameters of the fault detection model are obtained by training using historical equipment information of the switch cabinet, wherein the hyperparameters are parameters that need to be determined in the design phase of the fault detection model, and the internal parameters are the weights and biases of the fault detection model.

[0012] In one embodiment of the present application, a fault work order is generated based on the status of the switch cabinet, including: when the status of the switch cabinet is an insulation degradation state, a partial discharge state, or a temperature rise exceeding the standard state, a fault work order is generated based on the equipment information of the switch cabinet.

[0013] According to another aspect of an embodiment of the present invention, a fault handling device for a switch cabinet is also provided, including: an acquisition module, used to acquire equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulation material degradation gas information, equipment operating years and location information; a fault detection module, used to input the equipment information of the switch cabinet into a fault detection model, and perform fault detection on the switch cabinet based on the equipment information through the fault detection model to obtain the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm, and the state of the switch cabinet includes at least: normal state, insulation degradation state, partial discharge state and temperature rise exceeding standard state; a generation module, used to generate a fault work order according to the state of the switch cabinet; a processing module, used to perform fault handling on the switch cabinet based on the fault work order.

[0014] In an embodiment of the present invention, the equipment information of the switch cabinet is obtained, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material degradation gas information, equipment operating years and location information; the equipment information of the switch cabinet is input into the fault detection model, and the fault detection model is used to detect the fault of the switch cabinet based on the equipment information to obtain the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm; a fault work order is generated according to the state of the switch cabinet; and the switch cabinet is fault-handled based on the fault work order. It is easy to notice that the fault detection model obtained by optimizing the parameters based on the sparrow search algorithm is used to detect the fault of the switch cabinet, which improves the accuracy of the fault detection, and generates a fault work order according to the state of the switch cabinet, realizes the reasonable distribution of the fault information based on the fault work order, improves the rationality of the scheduling of the operation and maintenance resources when handling the fault, and achieves the technical effect of improving the fault handling efficiency of the switch cabinet, thereby solving the technical problem of low fault handling efficiency of the switch cabinet. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0016] Figure 1 is a flow chart of a switch cabinet fault processing method according to an embodiment of the present invention;

[0017] Figure 2 is a flow chart of an optional switch cabinet fault diagnosis and fault classification method according to an embodiment of the present invention;

[0018] Figure 3 is a schematic diagram of an optional fault handling device for a switch cabinet according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] Example 1

[0022] According to an embodiment of the present invention, an embodiment of a fault handling method for a switch cabinet is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0023] Figure 1 is a flow chart of a fault handling method of a switch cabinet according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0024] Step S102, obtaining equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material deterioration gas information, equipment operating years and location information.

[0025] The switchgear in the above steps is a device used to control, protect and distribute electric energy in the power system, and may include components such as circuit breakers, disconnectors, earthing switches, mutual inductors, busbars and cable terminals.

[0026] The device information in the above steps is used to characterize the status and characteristics of the switchgear.

[0027] The partial discharge information in the above steps may include but is not limited to the frequency, intensity and mode of partial discharge of the switch cabinet, which can be used to determine whether the insulation material of the switch cabinet has begun to deteriorate. Partial discharge is an electric field breakdown phenomenon that occurs in a local area of ​​the insulation system in power equipment.

[0028] The operating temperature information in the above steps may be temperature data inside the switch cabinet.

[0029] The operating humidity information in the above steps is used to characterize the humidity of the operating environment of the switch cabinet.

[0030] The insulating material degradation gas information in the above steps may be the type and concentration of gas released by the insulating material of the switch cabinet during the aging process, and may be used to discover potential insulation faults of the switch cabinet.

[0031] The equipment service life in the above steps is the length of time that has passed since the switchgear was first put into use, usually measured in years.

[0032] The location information in the above steps may be specific installation location information of the switch cabinet, which may include but is not limited to geographical coordinates, the location of the switch cabinet in the substation or distribution station, etc.

[0033] In an optional embodiment, different types of sensors may be deployed inside and around the switch cabinet to obtain equipment information. Specifically, an ultra-high frequency sensor may be deployed on the switch cabinet casing to capture the electromagnetic wave signal generated by the switch cabinet during the local discharge process to obtain local discharge information; a temperature sensor may be installed inside the switch cabinet to monitor the operating temperature inside the switch cabinet and key electrical components to obtain operating temperature information; a humidity sensor may be deployed in the surrounding environment of the switch cabinet to record the humidity of the operating environment of the switch cabinet to obtain operating humidity information; a gas detection sensor may be installed inside the switch cabinet to detect the gas generated by the decomposition of the insulating material to obtain gas information on deterioration of the insulating material.

[0034] At the same time, the location information of the switch cabinet can be obtained through the positioning system, and the equipment operating life of the switch cabinet can be determined based on the factory information of the switch cabinet.

[0035] After obtaining the equipment information of the switch cabinet, the obtained equipment information is cleaned to remove outliers and noise. Then the cleaned equipment information is normalized to unify the scale of the data. The formula for normalization is as follows:

[0036]

[0037] Among them, y is the value of the normalized data, y min is the minimum value of the normalized data range, y max is the maximum value of the data range after normalization, x is the value of the data before normalization, and x min is the minimum value of the data before normalization, x max is the maximum value of the data before normalization.

[0038] Step S104, input the equipment information of the switch cabinet into the fault detection model, perform fault detection on the switch cabinet based on the equipment information through the fault detection model, and obtain the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm, and the state of the switch cabinet includes at least: normal state, insulation degradation state, partial discharge state and temperature rise exceeding standard state.

[0039] The fault detection model in the above steps is used to identify whether the switch cabinet has a fault according to the equipment information of the switch cabinet, and can be an SSA-CNN (Sparrow Search Algorithm-Convolutional Neural Network) model, but is not limited to this.

[0040] The sparrow search algorithm in the above steps is a heuristic optimization algorithm. In the algorithm, the sparrow group is simulated as a set of solutions, and the best solution is found by searching in the solution space. In the fault detection model, the sparrow search algorithm can be used to optimize the parameters of the convolutional neural network.

[0041] The status of the switch cabinet in the above steps is used to quantify the failure risk of the switch cabinet.

[0042] The normal state in the above steps is used to indicate that the switch cabinet functions normally without obvious signs of failure.

[0043] The insulation degradation state in the above steps is used to indicate that the insulation material performance of the switch cabinet has declined and the risk of failure is high.

[0044] The partial discharge state in the above steps is used to characterize the situation where there is local power leakage in the insulation system of the equipment, and there is a risk of insulation system failure.

[0045] The temperature rise exceeding the limit state in the above steps is used to indicate that the operating temperature of the switch cabinet exceeds the safety threshold, which may be caused by overload, poor contact or heat dissipation system failure.

[0046] In an optional embodiment, an SSA-CNN model can be selected as a fault detection model, and the CNN (Convolutional Neural Network) part in the SSA-CNN model includes a convolution layer, a pooling layer, and a fully connected layer. The SSA-CNN model uses SSA (Sparrow Search Algorithm) to optimize the hyperparameters of the CNN part, that is, the parameters that need to be determined in the design stage, including the convolution kernel size, number, learning rate, etc., to improve the accuracy and efficiency of the fault detection model in diagnosing switch cabinet faults.

[0047] The convolution layer in CNN plays the role of feature extraction. The equipment information of the switch cabinet is used as the input matrix. After the input matrix is ​​input into the SSA-CNN model, the convolution kernel in the convolution layer is used to perform convolution operation on the equipment information of the switch cabinet to obtain the corresponding feature map. When the convolution kernel cannot reach the edge, the input matrix needs to be padded with zeros. The formula for the convolution operation is as follows:

[0048]

[0049] Among them, y is the feature map output by the convolution operation, ω is the input matrix, p is the number of zero-padding layers, k is the convolution kernel size, and s is the step size.

[0050] The pooling layer is located after the convolution layer, and the output of the convolution layer is taken as the input of the pooling layer. In the pooling layer, the maximum value or average value of the feature map is selected to replace other values ​​in the feature map, which reduces the amount of data and improves robustness. The fully connected layer is used to map the distributed features learned by the model to the sample label space. It is located after several convolution layers and pooling layers, and the fully connected layer outputs the state of the switch cabinet.

[0051] Exemplarily, an optional basic setting of the SSA-CNN model is as follows: the convolution layer is set to 2 layers, the convolution kernel size of each layer is 3×1, and the number is 16 and 32 respectively; the pooling layer and the batch normalization layer are set after the convolution layer; the activation function adopts the Re lu (Rectified Linear Unit) function, the fully connected layer adopts 2 layers, the number of neurons is 64 and 32 respectively, and after the softmax function, the output is 4; the optimizer adopts the Adam (Adaptive Moment Estimation) algorithm, and the learning rate is 0.001.

[0052] It should be noted that the above examples are only examples, and in fact the specific parameters of the fault detection model may not be limited thereto.

[0053] Step S106: Generate a fault work order according to the status of the switch cabinet.

[0054] The fault work order in the above steps is used to record the fault information of the switch cabinet, which may include but is not limited to the fault type, equipment location, fault discovery time, etc. The fault work order can be sent to the operation and maintenance personnel, who will repair the faulty switch cabinet recorded on the fault work order.

[0055] In an optional embodiment, whether a fault work order needs to be generated can be determined based on the status of the switch cabinet. When the status of the switch cabinet is abnormal, a fault work order can be created and information such as the fault type, detailed location of the switch cabinet, and time of fault discovery can be filled in the work order.

[0056] Step S108: Perform fault processing on the switch cabinet based on the fault work order.

[0057] In an optional embodiment, the fault work order may be randomly pushed to the operation and maintenance personnel so that the operation and maintenance personnel can respond in a timely manner and conduct on-site inspection or repair.

[0058] In another optional embodiment, the fault work orders may be classified first, and then pushed to the operation and maintenance personnel according to the classification results of the fault work orders. For example, the fault work orders may be classified according to the detailed location of the faulty switch cabinet, and then pushed to the operation and maintenance personnel responsible for the area; or the fault work orders may be classified according to the fault type, and then pushed to the operation and maintenance personnel responsible for the fault type.

[0059] In an embodiment of the present invention, the equipment information of the switch cabinet is obtained, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material degradation gas information, equipment operating years and location information; the equipment information of the switch cabinet is input into the fault detection model, and the fault detection model is used to detect the fault of the switch cabinet to obtain the state of the switch cabinet, wherein the fault detection model is a model obtained by optimizing the parameters of the convolutional neural network based on the sparrow search algorithm; a fault work order is generated according to the state of the switch cabinet; and the switch cabinet is fault-handled based on the fault work order. It is easy to notice that the fault detection model obtained by optimizing the parameters based on the sparrow search algorithm is used to detect the fault of the switch cabinet, which improves the accuracy of the fault detection, and generates a fault work order according to the state of the switch cabinet, realizes the reasonable distribution of the fault information based on the fault work order, improves the rationality of the scheduling of the operation and maintenance resources when handling the fault, and achieves the technical effect of improving the fault handling efficiency of the switch cabinet, thereby solving the technical problem of low fault handling efficiency of the switch cabinet.

[0060] In one implementation of the present application, fault processing of a switch cabinet based on a fault work order includes: classifying the fault work order to obtain a classification result of the fault work order; and performing fault processing on the switch cabinet based on the classification result.

[0061] In an optional embodiment, information such as the fault type, the location information of the switch cabinet, the environmental conditions under which the fault occurred, and the fault description can be extracted from the fault work order, and then the urgency of the fault is determined based on the extracted information, such as emergency repair, regular inspection, or preventive maintenance. Then, the fault work order is classified according to the urgency of the fault to obtain the classification result of the fault work order. Subsequently, the RPA (Robot ic Process Automation) robot is used to distribute the fault work order to different operation and maintenance personnel according to the classification result of the fault work order, such as distributing the fault work order requiring emergency repair to the maintenance personnel who handles the emergency task, and distributing the work order requiring regular inspection to the maintenance personnel who handles the regular inspection task. In the process of the operation and maintenance personnel performing fault handling on the switch cabinet according to the fault work order, the RPA robot continuously monitors the status of the switch cabinet and records the operation of the operation and maintenance personnel and the status change of the switch cabinet. After the fault handling is completed, the RPA robot collects the results of the fault handling, including whether the repair is successful, the specific cause of the fault, the measures taken, etc., and feeds the results of the fault handling back to the fault diagnosis model for the optimization of the fault diagnosis model.

[0062] In one implementation of the present application, fault work orders are classified to obtain classification results of the fault work orders, including: extracting features of the fault work orders to obtain a word vector matrix; inputting the word vector matrix into a classification model, and classifying the fault work orders based on the word vector matrix through the classification model to obtain classification results of the fault work orders.

[0063] The word vector matrix in the above steps is used to represent the mathematical structure of the text data in the fault ticket, and a fixed-length numerical vector can be used to represent the words in the fault ticket.

[0064] The classification model in the above steps is used to classify the fault work orders, and may be a CNN (convolutional neural network) model or a random forest model, but is not limited thereto.

[0065] In an optional embodiment, the words and phrases in the fault work order can be extracted from the fault work order features, and then the extracted words are vectorized to obtain a word vector matrix. The word vector matrix is ​​then input into the classification model, and a CNN model can be selected as the classification model. The CNN model is used as an example for explanation below.

[0066] Design the architecture of the CNN model, including convolutional layers, pooling layers, fully connected layers, etc., and determine the activation function and optimizer. Then use the historical word embedding matrix as the training data set to train the CNN model so that the trained CNN model has the ability to classify fault tickets according to the word embedding matrix. Then use the test data set to evaluate the CNN model, check the performance of the CNN model, and tune it according to the results to obtain a trained CNN model.

[0067] The word vector matrix is ​​input into the trained CNN model, and the CNN model classifies the fault work order according to the word vector matrix to obtain the classification result of the fault work order.

[0068] In one implementation of the present application, feature extraction is performed on a fault work order to obtain a word vector matrix, including: performing word segmentation processing on the fault work order to obtain a vocabulary list of the fault work order; inputting the vocabulary list into a word skipping model, and performing word vectorization processing on the vocabulary list through the word skipping model to obtain a word vector matrix.

[0069] The vocabulary in the above step is a collection of unique words extracted from the trouble tickets.

[0070] The word skipping model in the above steps is used to convert the words in the vocabulary into numerical vectors to form a word vector matrix.

[0071] In an optional embodiment, a word segmentation tool can be used to segment the text in the fault ticket to obtain a vocabulary of the fault ticket. The vocabulary is then input into the word skip model, where the construction process of the Skip-Gram model is as follows:

[0072] You can set the model parameters of the Skip-Gram model, such as the vector dimension, window size, and minimum word frequency. Then use the historical word vector matrix to train the Skip-Gram model through the stochastic gradient descent algorithm until convergence. This enables the Skip-Gram model to vectorize words in the vocabulary and obtain a trained Skip-Gram model.

[0073] The vocabulary is input into the Skip-Gram model. The Skip-Gram model counts the statistical relationship between the words in the vocabulary and the words that appear within a certain range around the word, and then vectorizes the statistical relationship. Each word in the vocabulary is mapped into a low-dimensional numerical vector, thereby generating a word vector matrix. The numerical vectors in the word vector matrix can be used as feature representations of fault tickets for subsequent classification tasks.

[0074] In one implementation of the present application, word segmentation is performed on a fault work order to obtain a vocabulary list of the fault work order, including: performing a data cleaning operation on the fault work order to obtain a fault work order after the data cleaning operation; performing word segmentation on the fault work order after the data cleaning operation to obtain data after word segmentation; performing deduplication processing on the data after word segmentation to obtain a vocabulary list of the fault work order.

[0075] The data cleaning process in the above steps is used to remove irrelevant characters in the fault work order.

[0076] The word segmentation process in the above steps is used to segment the continuous text sequence corresponding to the fault ticket into separate words.

[0077] The de-duplication process in the above steps is used to remove repeated words, and only one unique word is retained.

[0078] In an optional embodiment, the fault work order can be cleaned to remove irrelevant characters in the text of the fault work order, such as HTML (HyperText Markup Language) tags, special symbols, numbers, etc., to complete the redundancy removal of the data set and obtain the text data after data cleaning. Then, the jieba library (an open source library in Python) of Python (a programming language) is used to perform word segmentation on the text data after data cleaning to obtain the data after word segmentation. Then, the repeated words in the data after word segmentation are deleted, and only unique words are retained to obtain the vocabulary of the fault work order.

[0079] In one embodiment of the present application, the hyperparameters in the fault detection model are determined based on a sparrow search algorithm, and the internal parameters of the fault detection model are obtained by training using historical equipment information of the switch cabinet, wherein the hyperparameters are parameters that need to be determined in the design phase of the fault detection model, and the internal parameters are the weights and biases of the fault detection model.

[0080] The historical device information in the above steps is the device information of the switch cabinet during its historical operation.

[0081] The hyperparameters in the above steps are parameters set before the fault detection model is trained, and may include but are not limited to: learning rate, number of network layers, number of nodes in each layer, type of activation function, size and number of convolution kernels, etc.

[0082] The internal parameters in the above steps are parameters obtained by learning data during the fault detection model training process.

[0083] In an optional embodiment, before training the fault detection model, the hyperparameters in the fault detection model may be determined by a sparrow search algorithm.

[0084] In the sparrow search algorithm, according to the sparrow foraging process, the sparrow population is divided into three categories: discoverers, joiners, and vigilants. The sparrow search algorithm compares the fitness values ​​of the fault detection model and updates the positions of the three to find the optimal solution of the objective function, that is, the hyperparameters that are most suitable for the fault detection model.

[0085] The discoverer can propose new hyperparameter settings and evaluate their fitness through random exploration. During the iteration process, the discoverer's position is updated as shown in the formula:

[0086]

[0087] in, is the position of the i-th sparrow as the discoverer in the j-th dimension in the t-th iteration; t is the number of iterations; α is a random number, and the value range of α is a left-open and right-closed interval from 0 to 1; T max is the maximum number of iterations; Q is a random number that follows a normal distribution; L is an array of all 1s; R2 is the warning value; S T is a safe value.

[0088] When the warning value R2 is less than the safety value S T When the warning value R2 is greater than or equal to the safety value S T When , it indicates that there is a threat of predators in the environment, the sparrow population will fly to other places to forage.

[0089] During the foraging process, some joiners will observe the discoverer. When the discoverer finds better food, that is, a better hyperparameter combination, the joiners will compete with it, thereby updating the hyperparameter settings. The position update of the joiner is as follows:

[0090]

[0091] A + =A T (AA T ) -1 ;

[0092] in, is the position of the i-th sparrow in the j-th dimension as a joiner in the t-th iteration; Q is a random number that follows a normal distribution; X w is the global worst position; α is a random number, and the value range of α is a left-open and right-closed interval from 0 to 1; T max is the maximum number of iterations; is the global best position occupied by the discoverer in the t+1th iteration; X i,j is the position of the i-th sparrow in the j-th dimension; n is the total number of sparrows; A+ is the Moore-Penrose pseudo-inverse matrix of A, where A is an array, and the elements in the array are randomly 1 or -1. T is the transpose of A.

[0093] When i ≥ n / 2, it indicates that the i-th joiner with lower fitness has not obtained food and needs to go to another location.

[0094] When the sparrows are aware of danger, they will take anti-capture actions. If the sparrows find potential problems in the search process, such as the population falling into a local optimum, they will take action to change the positions of some individuals in the population and introduce new hyperparameter combinations. The position update formula of the sparrows is as follows:

[0095]

[0096] in, is the position of the i-th sparrow as a sentinel in the j-th dimension in the t-th iteration; is the global optimal solution in the tth iteration; β is a random number that conforms to the normal distribution with a mean of 0 and a variance of 1, which is used for step size control; K is a random number in [-1,1]; is the global worst position in the tth iteration; f i is the fitness value of the current sparrow i; f w is the current global worst fitness value; f b is the current global optimal fitness value.

[0097] When f i >f b When f i =f b , indicating that the sparrows in the middle of the population are aware of the danger and need to get closer to other sparrows to reduce the risk of being preyed upon.

[0098] The sparrow search algorithm continuously optimizes the hyperparameter settings in the population by iteratively executing the above behaviors. In each iteration, the performance of the individual is evaluated according to the fitness function, and the position of the individual in the population is updated until the preset number of iterations or performance convergence criteria are reached, and the hyperparameters in the fault detection model are obtained.

[0099] Then, an untrained fault detection model is constructed based on the hyperparameters, and the untrained fault detection model is trained using historical equipment information to determine the weights and biases of the fault detection model, that is, the internal parameters, so that the fault detection model has the ability to detect faults in the switchgear according to the equipment information.

[0100] In one embodiment of the present application, a fault work order is generated based on the status of the switch cabinet, including: when the status of the switch cabinet is an insulation degradation state, a partial discharge state, or a temperature rise exceeding the standard state, a fault work order is generated based on the equipment information of the switch cabinet.

[0101] In an optional embodiment, the insulation degradation state, partial discharge state, and temperature rise exceeding standard state in the switch cabinet are all abnormal states. In this state, the switch cabinet has a large potential fault. Therefore, a fault work order should be generated based on the equipment information so that maintenance personnel can conduct further inspection and maintenance on the switch cabinet in an abnormal state in time. When generating a fault work order, RPA (Robot ic Process Automation) technology can be used to use an RPA robot to generate a fault work order based on equipment information, but the method of generating a fault work order is not limited to this.

[0102] A preferred embodiment is described below. Figure 2 is a flow chart of an optional switch cabinet fault diagnosis and fault classification method according to an embodiment of the present invention, such as Figure 2 As shown:

[0103] Step S201, obtain switch cabinet fault sample data and divide it into a training set and a test set.

[0104] Step S202, initializing the sparrow population and calculating the initial position and fitness.

[0105] The sparrow search algorithm is used to determine the hyperparameters of the fault detection model. When using the sparrow search algorithm, it is necessary to first initialize the sparrow population and calculate the initial position and fitness.

[0106] Step S203, updating the sparrow population location.

[0107] Step S204, calculating the fitness of the sparrow population.

[0108] Step S205, determining whether the current number of iterations is greater than the maximum number of iterations.

[0109] The maximum number of iterations is preset. When the current number of iterations is greater than the maximum number of iterations, step S206 is continued to be executed; when the current number of iterations is less than or equal to the maximum number of iterations, step S203 is executed.

[0110] Step S206, outputting the hyperparameters of the fault detection model determined according to the sparrow search algorithm.

[0111] Step S207: establishing a fault detection model according to the hyperparameters.

[0112] Step S208: Use the training set and the test set to train and test the fault detection model.

[0113] Step S209: Use the trained fault detection model to perform fault diagnosis on the switch cabinet to obtain a fault diagnosis result.

[0114] The fault detection result is also the status of the switch cabinet mentioned above.

[0115] Step S210: Generate a fault work order based on the fault diagnosis result.

[0116] Step S211, establishing a classification model according to basic parameters.

[0117] A CNN model may be selected as a classification model, but is not limited thereto.

[0118] Step S212, training and testing the classification model.

[0119] Step S213: Use the trained classification model to classify the fault work order.

[0120] Example 2

[0121] According to an embodiment of the present invention, an embodiment of a fault handling device for a switch cabinet is provided. The device can execute the fault handling method for a switch cabinet provided in the above-mentioned embodiment 1. The specific implementation method and preferred application scenario are the same as those of the above-mentioned embodiment 1, and will not be repeated here.

[0122] Figure 3 is a schematic diagram of a fault handling device for a switch cabinet according to an embodiment of the present invention, such as Figure 3 As shown:

[0123] The acquisition module 30 is used to acquire the equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material deterioration gas information, equipment operating years and location information;

[0124] A fault detection module 32 is used to input the equipment information of the switch cabinet into the fault detection model, and perform fault detection on the switch cabinet based on the equipment information through the fault detection model to obtain the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm, and the state of the switch cabinet includes at least: normal state, insulation degradation state, partial discharge state and temperature rise exceeding standard state;

[0125] A generating module 34, used to generate a fault work order according to the status of the switch cabinet;

[0126] The processing module 36 is used to process the fault of the switch cabinet based on the fault work order.

[0127] The processing module includes: a classification unit, which is used to classify the fault work order and obtain the classification result of the fault work order; and a processing unit, which is used to perform fault processing on the switch cabinet based on the classification result.

[0128] The classification unit is also used to extract features from the fault work order to obtain a word vector matrix; the word vector matrix is ​​input into the classification model, and the fault work order is classified based on the word vector matrix through the classification model to obtain a classification result of the fault work order.

[0129] The classification unit is also used to perform word segmentation on the fault ticket to obtain a vocabulary list of the fault ticket; the vocabulary list is input into the word skipping model, and the vocabulary list is word vectorized through the word skipping model to obtain a word vector matrix.

[0130] The classification unit is also used to perform data cleaning operations on the fault work order to obtain the fault work order after the data cleaning operation; perform word segmentation processing on the fault work order after the data cleaning operation to obtain data after word segmentation processing; perform deduplication processing on the data after word segmentation processing to obtain a vocabulary list of the fault work order.

[0131] The hyperparameters in the fault detection model in the fault detection module are determined based on the sparrow search algorithm, and the internal parameters of the fault detection model are obtained by training using the historical equipment information of the switch cabinet. Among them, the hyperparameters are the parameters that need to be determined in the design phase of the fault detection model, and the internal parameters are the weights and biases of the fault detection model.

[0132] The generation module includes: a generation unit, which is used to generate a fault work order based on the equipment information of the switch cabinet when the state of the switch cabinet is an insulation degradation state, a partial discharge state, or a temperature rise exceeding the standard state.

[0133] Example 3

[0134] According to an embodiment of the present invention, there is further provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein the fault handling method for the switch cabinet in Embodiment 1 is executed when the program is running.

[0135] Example 4

[0136] An embodiment of the present application further provides a computer-readable storage medium, which includes a stored executable program, wherein when the executable program runs, the device where the computer-readable storage medium is located is controlled to execute the switch cabinet fault handling method in each embodiment of the present invention.

[0137] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0138] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0140] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0141] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0142] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0143] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A switch cabinet fault handling method, characterized in that: include: Acquire equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material deterioration gas information, equipment operating years and location information; Inputting the equipment information of the switch cabinet into a fault detection model, performing fault detection on the switch cabinet based on the equipment information through the fault detection model, and obtaining the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm, and the state of the switch cabinet includes at least: a normal state, an insulation degradation state, a partial discharge state, and a temperature rise exceeding the standard state; generating a fault work order according to the status of the switch cabinet; The switch cabinet is fault-handled based on the fault work order.

2. The switch cabinet fault handling method according to claim 1, characterized in that: Performing fault processing on the switch cabinet based on the fault work order includes: Classifying the fault work order to obtain a classification result of the fault work order; Fault processing is performed on the switch cabinet based on the classification result.

3. The switch cabinet fault handling method according to claim 2, characterized in that: Classifying the fault work order to obtain a classification result of the fault work order includes: Perform feature extraction on the fault work order to obtain a word vector matrix; The word vector matrix is ​​input into a classification model, and the fault work order is classified based on the word vector matrix by the classification model to obtain a classification result of the fault work order.

4. The switch cabinet fault handling method according to claim 3, characterized in that: Feature extraction is performed on the fault work order to obtain a word vector matrix, including: Performing word segmentation processing on the fault work order to obtain a vocabulary list of the fault work order; The vocabulary is input into a word skipping model, and the vocabulary is vectorized by the word skipping model to obtain the word vector matrix.

5. The switch cabinet fault handling method according to claim 4, characterized in that: Perform word segmentation processing on the fault work order to obtain a vocabulary list of the fault work order, including: Performing a data cleaning operation on the fault work order to obtain a fault work order after the data cleaning operation; Performing word segmentation processing on the fault work order after the data cleaning operation to obtain word segmentation processed data; The data after the word segmentation processing is deduplicated to obtain a vocabulary list of the fault work order.

6. The switch cabinet fault handling method according to claim 1, characterized in that: The hyperparameters in the fault detection model are determined based on the sparrow search algorithm, and the internal parameters of the fault detection model are obtained by training using historical equipment information of the switch cabinet, wherein the hyperparameters are parameters that need to be determined in the design phase of the fault detection model, and the internal parameters are weights and biases of the fault detection model.

7. The switch cabinet fault handling method according to claim 1, characterized in that: Generating a fault work order based on the state of the switch cabinet includes: When the state of the switch cabinet is the insulation degradation state, the partial discharge state, or the temperature rise exceeding the standard state, the fault work order is generated based on the equipment information of the switch cabinet.

8. A fault handling device for a switch cabinet, characterized in that: include: An acquisition module, used to acquire equipment information of the switch cabinet, wherein the equipment information includes at least one of the following: partial discharge information, operating temperature information, operating humidity information, insulating material deterioration gas information, equipment operating years and location information; A fault detection module, used for inputting the equipment information of the switch cabinet into a fault detection model, performing fault detection on the switch cabinet based on the equipment information through the fault detection model, and obtaining the state of the switch cabinet, wherein the fault detection model is a model obtained by adjusting the parameters of the convolutional neural network based on the sparrow search algorithm, and the state of the switch cabinet includes at least: normal state, insulation degradation state, partial discharge state and temperature rise exceeding standard state; A generation module, used to generate a fault work order according to the status of the switch cabinet; A processing module is used to perform fault processing on the switch cabinet based on the fault work order.

9. An electronic device, characterized in that: include: A memory storing an executable program; A processor is used to run the program, wherein the program executes the switch cabinet fault handling method according to any one of claims 1 to 7 when running.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is run, the device where the storage medium is located is controlled to execute the switch cabinet fault handling method according to any one of claims 1 to 7.

Images