Household fault intelligent identification method and system

Abstract

The invention discloses a household fault intelligent identification method and system, and relates to the technical field of household fault intelligent identification, and the method comprises the steps: collecting the power utilization data of a user, and carrying out the preprocessing; analyzing the preprocessed data by using an anomaly detection algorithm, and identifying an abnormal condition; extracting the preprocessed data through a deep learning algorithm to generate a high-dimensional feature vector; and when an abnormal condition is identified, early warning is carried out, the high-dimensional feature vector is analyzed, and a fault mode is identified. According to the method, through comprehensive calculation of various factors such as the path length and the feature weight, the combination condition of various abnormal features is accurately recognized, the anomaly detection precision is remarkably improved, and the problems of missing detection and false detection caused by simple threshold judgment in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of household faults, and in particular to a household fault intelligent identification method and system. Background Art

[0002] With the development of rural electrification, the residential electricity load is getting bigger and bigger, and the number of user fault reports is also increasing year by year. Most rural users do not understand electrical knowledge, do not have information such as electricity load, power outage maintenance plan information, and safe electricity use knowledge, and are unable to judge electrical fault modes and abnormal conditions. After a power outage at home, they can only call the power supply company to report a problem. The workload of the power supply company's operation and maintenance personnel has increased dramatically. Every year, the power supply company spends a lot of manpower, material resources, and financial resources on solving user faults. However, most of the reported faults are just simple air switch tripping or user power circuit leakage faults. Users only need troubleshooting guidance or troubleshooting steps to solve the problem by themselves.

[0003] In addition, existing power consumption anomaly detection often relies on simple threshold judgment or single feature analysis, which cannot accurately capture complex power consumption anomaly patterns, resulting in frequent missed detections and false detections. Due to the limitations of data processing and analysis algorithms, it is impossible to achieve real-time monitoring and rapid response to power consumption data, resulting in the inability to handle emergencies in a timely manner. In response to these shortcomings, we have invented a household fault intelligent identification method and system. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above problems or problems existing in the prior art, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is that due to the limitations of data processing and analysis algorithms, it is impossible to achieve real-time monitoring and rapid response to electricity consumption data, resulting in the inability to handle emergencies in a timely manner.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a household fault intelligent identification method, which includes collecting the user's electricity consumption data and preprocessing it;

[0008] Use anomaly detection algorithms to analyze preprocessed data and identify anomalies;

[0009] The preprocessed data is extracted through deep learning algorithms to generate high-dimensional feature vectors;

[0010] When an abnormal situation is identified, an early warning is issued and the high-dimensional feature vector is analyzed to identify the failure mode.

[0011] As a preferred solution of the household fault intelligent identification method described in the present invention, the collection of user electricity consumption data includes collecting user electrical parameters based on a household monitoring device, monitoring and realizing fault identification and protection actions such as overcurrent, short circuit, grounding, leakage, overvoltage and undervoltage, surge, etc., and transmitting the electrical parameter collection information and the protection switch action status to the concentrator; the concentrator uploads the concentrated information to the main station layer to store and analyze the data; the preprocessing includes data cleaning and data denoising.

[0012] As a preferred solution of the household fault intelligent identification method of the present invention, wherein: the deep learning algorithm includes:

[0013] The preprocessed electricity consumption data is subjected to feature extraction through convolutional neural network;

[0014] The extracted features are processed in time series through recurrent neural networks.

[0015] As a preferred solution of the household fault intelligent identification method of the present invention, wherein: the feature extraction includes:

[0016] Construct multiple convolutional layers to extract local features of data through convolution operations;

[0017] Use the pooling layer to reduce the dimension of the output of the convolutional layer, retain important features, and reduce the amount of calculation;

[0018] The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector for fault mode recognition.

[0019] As a preferred solution of the household fault intelligent identification method of the present invention, the extracting local features of the data includes extracting local features of the input data through multi-layer convolution operations, which is expressed as:

[0020]

[0021] Among them, f(τ) represents the input signal, g(t-τ) represents the convolution kernel, and τ is the integral variable;

[0022] The use of the pooling layer to reduce the dimension of the output of the convolutional layer includes maximum pooling, which is expressed as:

[0023] y i =max{u i,1 ,u i,2 ,…,u i,n}

[0024] Among them, u i,j represents the jth element in the i-th pooling area, y i Indicates the maximum value in the area;

[0025] The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector, which is expressed as:

[0026] z=W·u+b

[0027] Among them, W represents the weight matrix, u represents the output vector of the pooling operation, b represents the bias vector, and z is the generated high-dimensional feature vector.

[0028] As a preferred solution of the household fault intelligent identification method of the present invention, wherein: the anomaly detection algorithm includes establishing an improved isolation forest algorithm;

[0029] The improved isolation forest algorithm includes:

[0030] Build multiple decision trees, generated by randomly selecting features and split points;

[0031] Evaluate the abnormality of data points by calculating the path length of the data points in each decision tree;

[0032] Data points with high anomaly scores are marked as potential failure points.

[0033] As a preferred solution of the household fault intelligent identification method of the present invention, wherein: the path length of the calculated data point in each decision tree includes:

[0034] For each data point x, calculate its path length Li in each decision tree ,x ; The path length represents the number of edges that a data point passes through from the root node to the leaf node. The path length calculation formula is as follows:

[0035]

[0036] Among them, L i,x represents the path length of data point x in the i-th decision tree; d is the depth of the decision tree; 1 is the indicator function, which takes the value 1 when the condition is met, otherwise it takes the value 0; x fk represents the fth k eigenvalues; θ fk Indicates the f k The split point of the nodes;

[0037] The path length is adjusted by the exponential decay function and the path length adjustment coefficient to evaluate the abnormality of the data point, which is expressed as:

[0038]

[0039] Among them, λ is the exponential decay coefficient, which is used to control the rate of path length decay; t is the integral variable; γ is the adjustment coefficient of the path length, which is used to control the degree of influence of the path length;

[0040] The eigenvalues ​​are normalized so that different features have the same scale, expressed as:

[0041]

[0042] Among them, β ij is the normalization coefficient, which is used to control the normalization slope; x j is the jth eigenvalue of the data point; μ ij is the feature mean, which represents the mean of feature j in the i-th decision tree;

[0043] The adjusted path length is combined with the normalized eigenvalue and summed up, which is expressed as:

[0044]

[0045] in, represents the sum of all features j; W ij is the element of the weight matrix, representing the weight of the jth feature in the i-th decision tree;

[0046] The path length is integrated to evaluate the impact of path length on the anomaly score, expressed as:

[0047]

[0048] Calculate the path length of the data point x in each decision tree and evaluate its abnormality, expressed as:

[0049]

[0050] Among them, A(x) represents the abnormality degree of data point x; represents the sum of all decision trees;

[0051] The identifying of abnormal situations includes:

[0052] Collect and store users' historical electricity consumption data;

[0053] Compare the real-time collected electricity consumption data with historical data and calculate the anomaly score of the data point;

[0054] Abnormal electricity usage behavior is determined based on the abnormality score.

[0055] Another object of the present invention is to provide a system for intelligent identification of household faults, which can collect users' electricity consumption data and preprocess it; use an anomaly detection algorithm to analyze the preprocessed data and identify abnormal situations; extract the preprocessed data through a deep learning algorithm to generate a high-dimensional feature vector; when an abnormal situation is identified, issue an early warning and analyze the high-dimensional feature vector to identify the fault mode, thereby solving the problem of inaccurate intelligent identification of household faults.

[0056] As a preferred solution of the household fault intelligent identification system of the present invention, it includes: a processing module, an analysis module, a generation module, and an early warning module;

[0057] The processing and storage module is used to collect the user's electricity consumption data and perform pre-processing;

[0058] The analysis and prediction module is used to analyze the pre-processed data using an anomaly detection algorithm to identify anomalies;

[0059] The generation module is used to extract the preprocessed data through a deep learning algorithm to generate a high-dimensional feature vector;

[0060] The early warning module is used to issue an early warning when an abnormal situation is identified and to analyze the high-dimensional feature vector to identify the fault mode.

[0061] A computer-readable storage medium stores a computer program, which implements the steps of a household fault intelligent identification method when executed by a processor.

[0062] Beneficial effects of the present invention: The present invention conducts a comprehensive analysis of multi-dimensional electricity consumption data, and through comprehensive calculation of multiple factors such as path length and feature weight, accurately identifies the combination of multiple abnormal features, significantly improves the accuracy of abnormal detection, and solves the problems of missed detection and false detection caused by simple threshold judgment in the prior art. Through comprehensive analysis of multi-dimensional data, it is possible to identify complex abnormal electricity consumption patterns, solve the problem of single analysis dimension in the prior art, and can timely feedback abnormal information and provide specific processing suggestions, solve the problem of untimely user feedback in the prior art, and improve reliability and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0064] Figure 1The figure is the overall flow chart of the household fault intelligent identification method.

[0065] Figure 2 This is the overall flow chart of the household fault intelligent identification system.

[0066] Figure 3 A schematic diagram of the control hierarchy.

[0067] Figure 4 It is a schematic diagram of the overall device of the computer. DETAILED DESCRIPTION

[0068] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0069] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0070] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0071] Example 1

[0072] Reference Figures 1 to 4 , is an embodiment of the present invention, and provides a household fault intelligent identification method, comprising:

[0073] S1: Collect users’ electricity consumption data and perform preprocessing.

[0074] Furthermore, the collection of user electricity consumption data includes collecting user electrical parameters based on household monitoring devices, monitoring and implementing fault identification and protection actions such as overcurrent, short circuit, grounding, leakage, overvoltage and undervoltage, surge, etc., and transmitting the electrical parameter collection information and the protection switch action status to the concentrator; the concentrator uploads the concentrated information to the main station layer to manage data storage and analysis; the preprocessing includes data cleaning and data denoising.

[0075] S2: Use anomaly detection algorithms to analyze the preprocessed data and identify anomalies.

[0076] Further, the anomaly detection algorithm includes establishing an improved isolation forest algorithm;

[0077] The improved isolation forest algorithm includes:

[0078] Build multiple decision trees, generated by randomly selecting features and split points;

[0079] Evaluate the abnormality of data points by calculating the path length of the data points in each decision tree;

[0080] Data points with high anomaly scores are marked as potential failure points.

[0081] Furthermore, the calculation of the path length of the data point in each decision tree includes,

[0082] For each data point x, calculate its path length Li in each decision tree ,x ; The path length represents the number of edges that a data point passes through from the root node to the leaf node. The path length calculation formula is as follows:

[0083]

[0084] Among them, L i,x represents the path length of data point x in the i-th decision tree; d is the depth of the decision tree; 1 is the indicator function, which takes the value 1 when the condition is met, otherwise it takes the value 0; x fk represents the fth k eigenvalues; θ fk Indicates the f k The split point of the nodes;

[0085] The path length is adjusted by the exponential decay function and the path length adjustment coefficient to evaluate the abnormality of the data point, which is expressed as:

[0086]

[0087] Among them, λ is the exponential decay coefficient, which is used to control the rate of path length decay; t is the integral variable; γ is the adjustment coefficient of the path length, which is used to control the degree of influence of the path length;

[0088] The eigenvalues ​​are normalized so that different features have the same scale, expressed as:

[0089]

[0090] Among them, β ij is the normalization coefficient, which is used to control the normalization slope; x j is the jth eigenvalue of the data point; μ ij is the feature mean, which represents the mean of feature j in the i-th decision tree;

[0091] The adjusted path length is combined with the normalized eigenvalue and summed up, which is expressed as:

[0092]

[0093] in, represents the sum of all features j; W ij is the element of the weight matrix, representing the weight of the jth feature in the i-th decision tree;

[0094] The path length is integrated to evaluate the impact of path length on the anomaly score, expressed as:

[0095]

[0096] Calculate the path length of the data point x in each decision tree and evaluate its abnormality, expressed as:

[0097]

[0098] Among them, A(x) represents the abnormality degree of data point x; represents the sum of all decision trees, Li,x is the path length of data point x in the i-th decision tree, λ is the exponential decay coefficient of the path length, t is the integral variable, γ is the adjustment coefficient of the path length, Wij is the element of the weight matrix, βij is the normalization coefficient, xj is the j-th eigenvalue of the data point, μij is the eigenmean, e is a natural constant, represents the sum of all decision trees, represents the sum of all features, represents the integral from 0 to the path length Li,x.

[0099] It should be noted that the sum of the anomaly scores of all decision trees to obtain the total anomaly score of data point x is to comprehensively evaluate the degree of anomaly of the data point in the entire isolated forest; the score of a single decision tree may be affected by the randomly selected features and split points, resulting in unstable evaluation results. By summing the scores of multiple decision trees, the fluctuation of a single tree can be smoothed, and the stability and reliability of the evaluation results can be increased. Each decision tree will randomly select different features and split points during the construction process. Summing the scores of all decision trees can integrate the information of different features and split points to more comprehensively evaluate the degree of anomaly of the data point. A single decision tree may not be able to fully capture all the abnormal features of a data point. By summing the scores of multiple decision trees, more features and more split points can be used to enhance the accuracy of anomaly detection and reduce missed detections and false detections. The total anomaly score reflects the comprehensive performance of the data point in the entire isolated forest and quantifies the overall degree of anomaly of the data point. The higher the total anomaly score, the more likely the data point is an anomaly. By setting a suitable threshold, normal points and abnormal points can be effectively distinguished.

[0100] Furthermore, if the anomaly score A(x) of data point x is close to 0, the data point is recorded in the normal data log, and the user's electricity consumption data continues to be monitored in real time to ensure that the data remains normal;

[0101] When the anomaly score A(x) of a data point x is between 0 and 1, the data point is marked as a potential anomaly point and recorded in the potential anomaly log. The sampling frequency of the user's electricity consumption data is increased for more frequent monitoring. At the same time, the relevant information of the potential anomaly point is recorded in detail for subsequent analysis.

[0102] When the anomaly score A(x) of data point x is between 1 and 5, a suspected anomaly alarm is sent to the operation and maintenance personnel. After receiving the alarm, the operation and maintenance personnel conduct a detailed analysis of the data point, including checking the status of the power-consuming equipment, power consumption mode, etc., and notify the user to conduct inspection or maintenance if necessary based on the analysis results;

[0103] When the anomaly score A(x) of data point x is greater than 5, an emergency alarm is sent to the operation and maintenance center, and the data point is marked as an anomaly. After receiving the alarm, the operation and maintenance center immediately dispatches operation and maintenance personnel for on-site inspection. When the operation and maintenance personnel arrive at the site, they investigate the cause of the anomaly, check the electrical equipment and lines, and adjust parameters or perform equipment maintenance based on the on-site inspection results to ensure power safety. At the same time, they record the processing process and results of the anomaly and update the system log.

[0104] Furthermore, the identifying of abnormal situations includes:

[0105] Collect and store users' historical electricity consumption data;

[0106] Compare the real-time collected electricity consumption data with historical data and calculate the anomaly score of the data point;

[0107] Abnormal electricity usage behavior is determined based on the abnormality score.

[0108] S3: The preprocessed data is extracted through deep learning algorithms to generate high-dimensional feature vectors.

[0109] Furthermore, the deep learning algorithm includes:

[0110] The preprocessed electricity consumption data is subjected to feature extraction through convolutional neural network;

[0111] The extracted features are processed in time series through recurrent neural networks.

[0112] Furthermore, the feature extraction includes:

[0113] Construct multiple convolutional layers to extract local features of data through convolution operations;

[0114] Use the pooling layer to reduce the dimension of the output of the convolutional layer, retain important features, and reduce the amount of calculation;

[0115] The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector for fault mode recognition.

[0116] It should be noted that the extracting of local features of data includes extracting local features of input data through multi-layer convolution operations, which can be expressed as:

[0117]

[0118] Among them, f(τ) represents the input signal, g(t-τ) represents the convolution kernel, and τ is the integral variable;

[0119] The use of the pooling layer to reduce the dimension of the output of the convolutional layer includes maximum pooling, which is expressed as:

[0120] y i =max{u i,1 ,u i,2 ,…,u i,n}

[0121] Among them, u i,j represents the jth element in the i-th pooling area, y i Indicates the maximum value in the area;

[0122] The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector, which is expressed as:

[0123] z=W·u+b

[0124] Among them, W represents the weight matrix, u represents the output vector of the pooling operation, b represents the bias vector, and z is the generated high-dimensional feature vector.

[0125] S4: When an abnormal situation is identified, an early warning is issued and the high-dimensional feature vector is analyzed to identify the failure mode.

[0126] Furthermore, when an abnormal situation is identified, an early warning notification is sent through the user's terminal device (such as a mobile phone or tablet);

[0127] Analyze high-dimensional feature vectors;

[0128] Classify high-dimensional feature vectors through feature vector analysis algorithms to identify specific fault modes;

[0129] Generate corresponding fault handling solutions based on the identified fault modes.

[0130] In summary, the present invention conducts a comprehensive analysis of multi-dimensional electricity consumption data, and through the comprehensive calculation of multiple factors such as path length and feature weight, accurately identifies the combination of multiple abnormal features, significantly improves the accuracy of abnormal detection, and solves the problems of missed detection and false detection caused by simple threshold judgment in the prior art. Through the comprehensive analysis of multi-dimensional data, it is possible to identify complex abnormal electricity consumption patterns, solve the problem of single analysis dimension in the prior art, and can timely feedback abnormal information and provide specific processing suggestions, solve the problem of untimely user feedback in the prior art, and improve reliability and user satisfaction.

[0131] The above is a schematic scheme of a household fault intelligent identification method of this embodiment. It should be noted that the technical scheme of the household fault intelligent identification system and the technical scheme of the household fault intelligent identification method described above belong to the same concept, and the details of the technical scheme of the household fault intelligent identification system not described in detail in this embodiment can all be referred to the description of the technical scheme of the household fault intelligent identification method described above.

[0132] This embodiment also provides a household fault intelligent identification system, including:

[0133] Processing module, analysis module, generation module, early warning module;

[0134] The processing module is used to collect the user's electricity consumption data and perform pre-processing;

[0135] The analysis module is used to analyze the pre-processed data using an anomaly detection algorithm to identify anomalies;

[0136] The generation module is used to extract the preprocessed data through a deep learning algorithm to generate a high-dimensional feature vector;

[0137] The early warning module is used to issue an early warning when an abnormal situation is identified and to analyze the high-dimensional feature vector to identify the fault mode.

[0138] This embodiment further provides a computing device, which is applicable to the case of intelligent identification of household faults, including:

[0139] Memory and processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the household fault intelligent identification method proposed in the above embodiment.

[0140] This embodiment further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for intelligently identifying household faults as proposed in the above embodiment is implemented.

[0141] The storage medium proposed in this embodiment and the method for realizing intelligent identification of household faults proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0142] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.

[0143] If the function 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, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0144] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied 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 the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0145] 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 processing in another suitable manner as necessary, and then stored in a computer memory.

[0146] 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 their combination: 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. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0147] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A household fault intelligent identification method, characterized by: include, Collect users' electricity consumption data and pre-process them; Use anomaly detection algorithms to analyze preprocessed data and identify anomalies; The preprocessed data is extracted through deep learning algorithms to generate high-dimensional feature vectors; When an abnormal situation is identified, an early warning is issued and the high-dimensional feature vector is analyzed to identify the failure mode.

2. The household fault intelligent identification method according to claim 1, characterized in that: The collecting of the user's electricity consumption data includes collecting the user's electrical parameters based on the household monitoring device, monitoring and realizing fault identification and protection actions such as overcurrent, short circuit, grounding, leakage, overvoltage, undervoltage, surge, etc., and transmitting the electrical parameter collection information and the protection switch action status to the concentrator; The concentrator uploads the concentrated information to the main station layer for data storage and analysis management; the preprocessing includes data cleaning and data denoising.

3. The household fault intelligent identification method according to claim 2, characterized in that: The deep learning algorithm includes: The preprocessed electricity consumption data is subjected to feature extraction through convolutional neural network; The extracted features are processed in time series through recurrent neural networks.

4. The household fault intelligent identification method according to claim 3, characterized in that: The feature extraction comprises: Construct multiple convolutional layers to extract local features of data through convolution operations; Use the pooling layer to reduce the dimensionality of the output of the convolutional layer; The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector for fault mode recognition.

5. The household fault intelligent identification method according to claim 4, characterized in that: The extracting of local features of the data includes extracting local features of the input data through multi-layer convolution operations, which is expressed as: Among them, f(τ) represents the input signal, g(t-τ) represents the convolution kernel, and τ is the integral variable; The use of the pooling layer to reduce the dimension of the output of the convolutional layer includes maximum pooling, which is expressed as: y i =max{u i,1 ,u i,2 ,…,u i，n } Among them, u i,j represents the jth element in the i-th pooling area, y i Indicates the maximum value in the area; The output of the pooling layer is passed to the fully connected layer to generate a high-dimensional feature vector, which is expressed as: z=W·u+b Among them, W represents the weight matrix, u represents the output vector of the pooling operation, b represents the bias vector, and z is the generated high-dimensional feature vector.

6. The household fault intelligent identification method according to claim 5, characterized in that: The anomaly detection algorithm includes establishing an improved isolation forest algorithm; The improved isolation forest algorithm includes: Build multiple decision trees, generated by randomly selecting features and split points; Evaluate the abnormality of data points by calculating the path length of the data points in each decision tree; Data points with high anomaly scores are marked as potential failure points.

7. The household fault intelligent identification method according to claim 6, characterized in that: The calculation of the path length of the data point in each decision tree includes: For each data point x, calculate its path length Li in each decision tree ,x ; The path length represents the number of edges that a data point passes through from the root node to the leaf node. The path length calculation formula is as follows: Among them, L i,x represents the path length of data point x in the i-th decision tree; d is the depth of the decision tree; 1 is the indicator function; x fk represents the fth k eigenvalues; θ fk Indicates the f k The split point of the nodes; The path length is adjusted by the exponential decay function and the path length adjustment coefficient to evaluate the abnormality of the data point, which is expressed as: Among them, λ is the exponential decay coefficient; t is the integral variable; γ is the adjustment coefficient of the path length; The eigenvalues ​​are normalized so that different features have the same scale, expressed as: Among them, β ij is the normalization coefficient; x j is the jth eigenvalue of the data point; μ ij is the feature mean; The adjusted path length is combined with the normalized eigenvalue and summed up, which is expressed as: in, represents the sum of all features j; W ij are the elements of the weight matrix; The path length is integrated to evaluate the impact of path length on the anomaly score, expressed as: Calculate the path length of the data point x in each decision tree and evaluate its abnormality, expressed as: Among them, A(x) represents the abnormality degree of data point x; represents the sum of all decision trees; The identifying of abnormal situations includes: Collect and store users' historical electricity consumption data; Compare the real-time collected electricity consumption data with historical data and calculate the anomaly score of the data point; Abnormal electricity usage behavior is determined based on the abnormality score.

8. A system using the household fault intelligent identification method according to any one of claims 1 to 7, characterized in that: It includes processing module, analysis module, generation module and early warning module; The processing module is used to collect the user's electricity consumption data and perform pre-processing; The analysis module is used to analyze the pre-processed data using an anomaly detection algorithm to identify anomalies; The generation module is used to extract the preprocessed data through a deep learning algorithm to generate a high-dimensional feature vector; The early warning module is used to issue an early warning when an abnormal situation is identified and to analyze the high-dimensional feature vector to identify the fault mode.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the household fault intelligent identification method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the household fault intelligent identification method described in any one of claims 1 to 7 are implemented.

Images