Electrical management method and system based on artificial intelligence

Abstract

The invention discloses an electrical management method and system based on artificial intelligence, and belongs to the technical field of electrical management, and the method comprises the steps: carrying out the discretization of a preprocessed historical state value, obtaining a discretized historical state value, carrying out the correlation support degree analysis of the discretized historical state value, determining a correlation support degree analysis result, and carrying out the analysis of the correlation support degree. And on the basis of the association support degree analysis result, determining a target influence factor corresponding to each fault type, learning a historical state value corresponding to the target influence factor by adopting artificial intelligence to obtain a fault identification rule, and finally managing the electrical equipment through the fault identification rule. The management efficiency can be effectively improved, faults can be found in time, a more effective basis is provided for maintenance, and the electrical management effect is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of electrical management, and particularly relates to an electrical management method and system based on artificial intelligence. Background Art

[0002] Electrical equipment refers to devices and systems used for generating, transmitting, distributing, converting, and consuming electrical energy. They play a crucial role in modern industry and daily life. With the rapid development of China's economy, the demand for electricity is continuously increasing, and the scale and complexity of electrical equipment are also constantly improving. Traditional electrical management methods mainly rely on manual inspections and regular maintenance, and have the following problems: low management efficiency, high labor costs; failure to detect faults in a timely manner, which is likely to cause safety accidents; lack of scientific basis for maintenance decisions, which may lead to over-maintenance or under-maintenance. Summary of the Invention

[0003] The present invention provides an electrical management method and system based on artificial intelligence to solve the problems of low management efficiency, failure to detect faults in a timely manner, and lack of scientific basis for maintenance decisions existing in the prior art.

[0004] On the one hand, the present invention provides an electrical management method based on artificial intelligence, including: For any one type of fault of electrical equipment, collect the historical state values corresponding to all influencing factors under the fault type, and preprocess the historical state values to obtain the preprocessed historical state values; Discretize the preprocessed historical state values to obtain the discretized historical state values, and perform association support degree analysis on the discretized historical state values to determine the association support degree analysis result; Based on the association support degree analysis result, determine the target influencing factors corresponding to each type of fault, and use artificial intelligence to learn the historical state values corresponding to the target influencing factors to obtain fault recognition rules; Collect the real-time state values corresponding to the target influencing factors, and use the fault recognition rules to identify the real-time state values, determine the fault diagnosis result of the electrical equipment, and give an early warning according to the fault diagnosis result to complete electrical management.

[0005] Further, for any one type of fault of electrical equipment, collecting the historical state values corresponding to all influencing factors under the fault type, and preprocessing the historical state values to obtain the preprocessed historical state values, includes: For any one type of fault of electrical equipment, when the data sampling period arrives, collect the historical state values corresponding to all influencing factors under the fault type to obtain the historical state values corresponding to each type of fault; Normalize the historical state values corresponding to each fault type to obtain the historical state values after preprocessing.

[0006] Furthermore, all influencing factors include the temperature change rate, ambient temperature, voltage, overvoltage rate of voltage, current, overcurrent rate of current, internal resistance of the electrical equipment, and the change rate of the internal resistance of the electrical equipment.

[0007] Furthermore, discretize the historical state values after preprocessing to obtain the historical state values after discretization, and perform an association support degree analysis on the historical state values after discretization to determine the association support degree analysis results, including: For the historical state values after preprocessing, pair the influencing factors two by two to obtain multiple influencing factor pairing combinations, and form coordinate points with the historical state values corresponding to the influencing factor pairing combinations to obtain the historical state values after discretization; For the historical state values after discretization in the discrete space, use the K-means clustering algorithm for clustering to determine the clustering results; among them, the number of K-means clustering results is the same as the number of types of fault types; For any one type of fault, based on the clustering results, determine the association support degree of the influencing factor pairing combination to the fault type to obtain the association support degree analysis results.

[0008] Furthermore, for the historical state values after preprocessing, pair the influencing factors two by two to obtain multiple influencing factor pairing combinations, and form coordinate points with the historical state values corresponding to the influencing factor pairing combinations to obtain the historical state values after discretization, including: Put all the influencing factors into a set, randomly select one influencing factor, and pair the selected influencing factor with other influencing factors in the set two by two to obtain influencing factor pairing combinations; Judge whether the remaining influencing factors in the set are zero. If so, complete the pairwise pairing of the influencing factors to obtain the influencing factor pairing combinations corresponding to each selected influencing factor. Otherwise, return to the step of obtaining the influencing factor pairing combinations; For any one influencing factor pairing combination, use the historical state values at any one historical moment corresponding to the influencing factor pairing combination as the abscissa point and the ordinate point respectively to obtain the historical state values after discretization.

[0009] Furthermore, for the historical state values after discretization in the discrete space, use the K-means clustering algorithm for clustering to determine the clustering results, including: Based on the number of types of fault types K, randomly select K clustering centers; where K is a positive integer; For the historical state values after discretization in the discrete space, obtain the Euclidean distance between each coordinate point and the cluster center, and classify the coordinate points into the cluster centers according to the Euclidean distance; After all coordinate points are classified, calculate the average values of all data in the K classes respectively, obtain K average values and use them as the new cluster centers; Judge whether the cluster centers do not change continuously for multiple times. If so, output the clustering result; otherwise, return to the step of classifying the coordinate points into the cluster centers.

[0010] Furthermore, for any type of fault, based on the clustering result, determine the associated support degree of the influencing factor pairing combination to the fault type, and obtain the associated support degree analysis result, including: For any type of fault, based on the clustering result, determine the number of coordinate points regarding the fault type in each cluster category; According to the number of coordinate points regarding the fault type in each cluster category, determine that the fault type with the largest number of coordinate points is the true fault type corresponding to the cluster category; For any type of fault, determine that the cluster categories with the same true fault type and the largest number of coordinate points are the target cluster categories; For the target cluster category, determine the proportion of the number of coordinate points determined as the true fault type in the target cluster category, and obtain the associated support degree analysis result.

[0011] Furthermore, based on the associated support degree analysis result, determine the target influencing factors corresponding to each type of fault, and use artificial intelligence to learn the historical state values corresponding to the target influencing factors to obtain the fault recognition rules, including: Judge whether the associated support degree analysis result is greater than the preset support degree threshold. If so, use the influencing factor corresponding to the coordinate point as the target influencing factor corresponding to the true fault type; otherwise, do not use the influencing factor corresponding to the coordinate point as the target influencing factor corresponding to the true fault type; Obtain the target influencing factors corresponding to each type of fault, use the historical state value of the target influencing factor at any historical moment as the training data, and use the corresponding fault type as the training label; Initialize the fault recognition rules with a neural network, and use the training data and training labels to train the fault recognition rules to obtain the trained fault recognition rules.

[0012] Furthermore, collect the real-time state values corresponding to the target influencing factors, use the fault recognition rules to identify the real-time state values, determine the fault diagnosis result of the electrical equipment, and give an early warning according to the fault diagnosis result, including: Collect the real-time status values corresponding to the target influencing factors, form the real-time status values corresponding to the target influencing factors into the input of the fault identification rule, and perform identification through the fault identification rule to obtain the fault diagnosis result; When the fault diagnosis result is of an abnormal category, a text warning message is generated and the text warning message is transmitted to the device designated by the staff to achieve warning.

[0013] On the other hand, the present invention provides an electrical management system based on artificial intelligence, including: a data acquisition module, a data analysis module, an artificial intelligence model, and an electrical management module; The data acquisition module is used to collect the historical status values corresponding to all influencing factors under a fault type for any fault type of electrical equipment, and preprocess the historical status values to obtain the historical status values after preprocessing; The data analysis module is used to discretize the historical status values after preprocessing to obtain the historical status values after discretization, and perform association support degree analysis on the historical status values after discretization to determine the association support degree analysis result; The artificial intelligence model is used to determine the target influencing factors corresponding to each fault type based on the association support degree analysis result, and use artificial intelligence to learn the historical status values corresponding to the target influencing factors to obtain the fault identification rule; The electrical management module is used to collect the real-time status values corresponding to the target influencing factors, identify the real-time status values using the fault identification rule, determine the fault diagnosis result of the electrical equipment, and perform warning according to the fault diagnosis result to complete electrical management.

[0014] An electrical management method based on artificial intelligence provided by the present invention discretizes the historical status values after preprocessing to obtain the historical status values after discretization, performs association support degree analysis on the historical status values after discretization to determine the association support degree analysis result, determines the target influencing factors corresponding to each fault type based on the association support degree analysis result, and uses artificial intelligence to learn the historical status values corresponding to the target influencing factors to obtain the fault identification rule. Finally, the electrical equipment is managed through the fault identification rule, which can effectively improve the management efficiency, timely detect faults, and provide a more effective basis for maintenance, greatly improving the electrical management effect. Description of the Drawings

[0015] The drawings here are incorporated into the specification and form a part of this specification, showing the embodiments in line with the present invention, and are used together with the specification to explain the principles of the present invention.

[0016] Figure 1Flowchart of an electrical management method based on artificial intelligence provided by an embodiment of the present invention.

[0017] Figure 2 Schematic structural diagram of an electrical management system based on artificial intelligence provided by an embodiment of the present invention.

[0018] Among them, 201 - data acquisition module, 202 - data analysis module, 203 - artificial intelligence model, 204 - electrical management module.

[0019] Through the above-mentioned drawings, specific embodiments of the present invention have been shown, and there will be more detailed descriptions hereinafter. These drawings and text descriptions are not intended to limit the scope of the inventive concept in any way, but to illustrate the concept of the present invention to those skilled in the art by referring to specific embodiments. Detailed implementation manners

[0020] Here, exemplary embodiments will be described in detail, and their examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present invention. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0021] The embodiments of the present invention will be described in detail below with reference to the drawings.

[0022] As Figure 1 shown, an embodiment of the present invention provides an electrical management method based on artificial intelligence, including: S101. For any one type of fault of an electrical device, collect the historical state values corresponding to all influencing factors under the fault type, and preprocess the historical state values to obtain the preprocessed historical state values; For example, assume that the electrical device is a power battery, and its fault types may include abnormal battery pack temperature, abnormal charging, abnormal discharging, etc., and the influencing factors may include temperature change rate, ambient temperature, voltage, overvoltage rate of voltage, current, overcurrent rate of current, internal resistance of the electrical device, and change rate of the internal resistance of the electrical device. Then, the historical state values corresponding to all influencing factors under normal, abnormal battery pack temperature, abnormal charging, abnormal discharging and other working conditions should be collected, and then the historical state values are preprocessed to facilitate data processing. It should be noted that the above-mentioned electrical device, fault type, and influencing factors are all examples and can be selected according to actual situations.

[0023] In the prior art, after preprocessing all influencing factors, artificial intelligence is often directly used for data learning and data recognition. Therefore, the correlation support degree of data for fault types is ignored. A large amount of irrelevant data will lead to low recognition accuracy and ultimately reduce the management efficiency of electrical equipment. Therefore, the present invention will next analyze the correlation support degree to determine some influencing factors with a strong correlation with the fault type, so as to improve the fault recognition efficiency and accuracy.

[0024] S102. Discretize the historical state values after preprocessing to obtain the discretized historical state values, and perform correlation support degree analysis on the discretized historical state values to determine the correlation support degree analysis result. The clustering algorithm can be used to discretize the historical state values after preprocessing, thereby transforming the numerical time series into a symbolic sequence. Then, the method of correlation support degree analysis can be used to discover the correlation therein, and finally determine the correlation support degree of the influencing factor combination state for fault recognition, so as to facilitate the analysis of the influencing factors.

[0025] S103. Based on the correlation support degree analysis result, determine the target influencing factors corresponding to each fault type, and use artificial intelligence to learn the historical state values corresponding to the target influencing factors to obtain fault recognition rules. A correlation support degree threshold can be preset in advance. When the correlation support degree in the correlation support degree analysis result is greater than this correlation support degree threshold, it can be considered that the corresponding influencing factor supports the recognition of this fault type. Therefore, after determining the target influencing factors corresponding to each fault type, for any fault type, after training the artificial intelligence with the historical state values corresponding to the target influencing factors, the artificial intelligence after training can be used to recognize this fault type, so as to achieve more accurate fault recognition and effectively improve the management efficiency of electrical equipment.

[0026] S104. Collect the real-time state values corresponding to the target influencing factors, and use the fault recognition rules to recognize the real-time state values to determine the fault diagnosis result of the electrical equipment, and give an early warning according to the fault diagnosis result to complete electrical management.

[0027] Optionally, giving an early warning according to the fault diagnosis result may include warning methods such as sound and light warning and information warning, so as to timely remind the staff to manage the electrical equipment.

[0028] An electrical management method based on artificial intelligence provided by the present invention discretizes the historical state values after preprocessing to obtain the discretized historical state values, performs an association support degree analysis on the discretized historical state values to determine the association support degree analysis result, determines the target influencing factors corresponding to each fault type based on the association support degree analysis result, and uses artificial intelligence to learn the historical state values corresponding to the target influencing factors to obtain fault recognition rules. Finally, the electrical equipment is managed through the fault recognition rules, which can effectively improve the management efficiency, timely detect faults, and provide a more effective basis for maintenance, greatly enhancing the electrical management effect.

[0029] Compared with the prior art, the technical solution provided by the embodiments of the present invention can analyze the target influencing factors corresponding to various faults regardless of the type of electrical equipment, thereby improving the fault recognition accuracy.

[0030] In the embodiments of the present invention, for any fault type of electrical equipment, the historical state values corresponding to all influencing factors under the fault type are collected, and the historical state values are preprocessed to obtain the preprocessed historical state values, including: For any fault type of electrical equipment, when the data sampling period arrives, the historical state values corresponding to all influencing factors under the fault type are collected to obtain the historical state values corresponding to each fault type; The historical state values corresponding to each fault type are normalized to obtain the preprocessed historical state values.

[0031] In the embodiments of the present invention, by setting the data sampling period, every other period, the fault recognition accuracy can be gradually improved according to the actual data, and the traditional fault inspection of electrical equipment can be gradually converted into automatic inspection. As time goes by, the fault recognition accuracy gradually rises and stabilizes, that is, automatic inspection can be achieved.

[0032] In the embodiments of the present invention, all influencing factors include the temperature change rate, ambient temperature, voltage, voltage overvoltage rate, current, current overcurrent rate, internal resistance of the electrical equipment, and internal resistance change rate of the electrical equipment.

[0033] It should be noted that the above influencing factors are only some relatively simple examples, and other influencing factors can also be used for analysis to finally determine the target influencing factors that are most conducive to recognition.

[0034] In the embodiments of the present invention, the preprocessed historical state values are discretized to obtain the discretized historical state values, and an association support degree analysis is performed on the discretized historical state values to determine the association support degree analysis result, including: For the historical state values after preprocessing, pair the influencing factors pairwise to obtain multiple paired combinations of influencing factors, and form coordinate points with the historical state values corresponding to the paired combinations of influencing factors to obtain the historical state values after discretization. For the historical state values after discretization in the discrete space, perform clustering processing using the K-means clustering algorithm to determine the clustering results; among them, the number of K-means clustering results is the same as the number of types of fault types. For any one type of fault, based on the clustering results, determine the association support degree of the paired combination of influencing factors for the fault type to obtain the association support degree analysis result.

[0035] In the embodiment of the present invention, for the historical state values after preprocessing, pair the influencing factors pairwise to obtain multiple paired combinations of influencing factors, and form coordinate points with the historical state values corresponding to the paired combinations of influencing factors to obtain the historical state values after discretization, including: Put all the influencing factors into a set, randomly select one influencing factor, and pair the selected influencing factor pairwise with other influencing factors in the set to obtain paired combinations of influencing factors. Judge whether the remaining influencing factors in the set are zero. If so, complete the pairwise pairing of the influencing factors to obtain the paired combinations of influencing factors corresponding to each selected influencing factor. Otherwise, return to the step of obtaining the paired combinations of influencing factors. For any paired combination of influencing factors, use the historical state values at any historical moment corresponding to the paired combination of influencing factors as the abscissa point and the ordinate point respectively to obtain the historical state values after discretization.

[0036] The clustering algorithm can be used to discretize the historical state values after preprocessing, thereby transforming the numerical time series into a symbolic sequence. Then, the method of association support degree analysis can be used to discover the correlation therein, and finally determine the association support degree of the influencing factor combination state for fault identification to facilitate the analysis of the influencing factors.

[0037] In the embodiment of the present invention, for the historical state values after discretization in the discrete space, perform clustering processing using the K-means clustering algorithm to determine the clustering results, including: Based on the number K of types of fault types, randomly select K clustering centers; where K is a positive integer. For the historical state values after discretization in the discrete space, obtain the Euclidean distance between each coordinate point and the clustering center, and classify the coordinate points into the clustering center according to the Euclidean distance, that is, the coordinate points are classified into the nearest clustering center.

[0038] After all coordinate points are classified, calculate the average values of all data in the K classes respectively, obtain K average values and use them as the new clustering centers; Determine whether the clustering centers do not change for multiple consecutive times. If so, output the clustering result; otherwise, return to the step of classifying the coordinate points to the clustering centers.

[0039] Optionally, in addition to using the K-means clustering algorithm for clustering, other clustering algorithms can also be used for clustering, which can also achieve the object of the present invention.

[0040] In the embodiment of the present invention, for any type of fault, based on the clustering result, determine the association support degree of the influence factor pairing combination on the fault type, and obtain the association support degree analysis result, including: For any type of fault, based on the clustering result, determine the number of coordinate points regarding the fault type in each clustering category; According to the number of coordinate points regarding the fault type in each clustering category, determine that the fault type with the largest number of coordinate points is the true fault type corresponding to the clustering category; For any type of fault, determine that the clustering categories with the same true fault type and the largest number of coordinate points are the target clustering categories; For the target clustering category, determine the proportion of the number of coordinate points determined as the true fault type in the target clustering category, and obtain the association support degree analysis result.

[0041] For example, there are three types of faults, a, b, and c. For any influence factor pairing combination, after clustering the influence factor pairing combination, three clustering categories, A, B, and C, are obtained. The most common fault type corresponding to the influence factor pairing combination in A is a, then a is the target clustering category. In actual situations, the most common fault type corresponding to the influence factor pairing combination in B may also be a. Then, it is determined whether there are more a in A or more a in B, and the larger proportion is used as the association support degree analysis result (assuming that there are more a in B, and the number of a is L1, and the number of influence factor pairing combinations in B is L2, that is, the number of coordinate points is L2, then the association support degree analysis result is L1 / L2).

[0042] In the embodiment of the present invention, based on the association support degree analysis result, determine the target influence factor corresponding to each fault type, and use artificial intelligence to learn the historical state values corresponding to the target influence factor to obtain the fault recognition rule, including: Determine whether the association support degree analysis result is greater than the preset support degree threshold. If so, use the influence factor corresponding to the coordinate point as the target influence factor corresponding to the true fault type; otherwise, do not use the influence factor corresponding to the coordinate point as the target influence factor corresponding to the true fault type; Obtain the target influencing factors corresponding to each fault type, and use the historical state values of the target influencing factors at any historical moment as training data, and use the corresponding fault type as the training label; Initialize the fault recognition rule using a neural network, and train the fault recognition rule using the training data and the training label to obtain the trained fault recognition rule.

[0043] It should be noted that since the target influencing factors corresponding to each fault type are mostly different, therefore, it is necessary to train the fault recognition rule separately for each fault type, and detect the fault type through the fault recognition rule, so as to achieve accurate detection.

[0044] In the embodiment of the present invention, collect the real-time state values corresponding to the target influencing factors, and use the fault recognition rule to identify the real-time state values, determine the fault diagnosis result of the electrical equipment, and give an early warning according to the fault diagnosis result, including: Collect the real-time state values corresponding to the target influencing factors, form the input of the fault recognition rule with the real-time state values corresponding to the target influencing factors, and perform recognition through the fault recognition rule to obtain the fault diagnosis result; When the fault diagnosis result is an abnormal category, generate a text warning message and transmit the text warning message to the device designated by the staff to achieve early warning.

[0045] As Figure 2 shown, the present invention provides an electrical management system based on artificial intelligence, including: a data acquisition module 201, a data analysis module 202, an artificial intelligence model 203, and an electrical management module 204; The data acquisition module 201 is used to collect the historical state values corresponding to all influencing factors under a fault type for any fault type of the electrical equipment, and preprocess the historical state values to obtain the preprocessed historical state values; The data analysis module 202 is used to discretize the preprocessed historical state values to obtain the discretized historical state values, and perform association support degree analysis on the discretized historical state values to determine the association support degree analysis result; The artificial intelligence model 203 is used to determine the target influencing factors corresponding to each fault type based on the association support degree analysis result, and use artificial intelligence to learn the historical state values corresponding to the target influencing factors to obtain a fault recognition rule; The electrical management module 204 is used to collect the real-time state values corresponding to the target influencing factors, and use the fault recognition rule to identify the real-time state values, determine the fault diagnosis result of the electrical equipment, and give an early warning according to the fault diagnosis result to complete electrical management.

[0046] The electrical management system based on artificial intelligence provided by the embodiments of the present invention can execute the above method technical solution, and its principle and beneficial effects are similar, so details are not described herein again.

[0047] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing in the process Figure 1 one process or multiple processes and / or blocks Figure 1 a device for the functions specified in one block or multiple blocks.

[0049] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes in the process Figure 1 one process or multiple processes and / or blocks Figure 1 a device for the functions specified in one block or multiple blocks.

[0050] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 a device for the functions specified in one block or multiple blocks.

[0051] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above facts and methods can be completed by instructing relevant hardware through a program. The program involved or the said program can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: At this time, the corresponding method steps are introduced. The said storage medium can be ROM / RAM, magnetic disk, optical disk, etc.

[0052] The specific embodiments described above have further elaborated on the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An electrical management method based on artificial intelligence, characterized in that: include: For any fault type of the electrical equipment, collect the historical state values ​​corresponding to all influencing factors under the fault type, and pre-process the historical state values ​​to obtain the historical state values ​​after pre-processing; Discretizing the historical state values ​​after preprocessing to obtain the discretized historical state values, and performing an association support analysis on the discretized historical state values ​​to determine the association support analysis results; Based on the correlation support analysis results, the target influencing factors corresponding to each fault type are determined, and artificial intelligence is used to learn the historical state values ​​corresponding to the target influencing factors to obtain fault identification rules; Collect the real-time status values ​​corresponding to the target influencing factors, and use fault identification rules to identify the real-time status values, determine the fault diagnosis results of the electrical equipment, and issue early warnings based on the fault diagnosis results to complete electrical management.

2. The electrical management method based on artificial intelligence according to claim 1, characterized in that: For any fault type of the electrical equipment, historical status values ​​corresponding to all influencing factors under the fault type are collected, and the historical status values ​​are preprocessed to obtain the preprocessed historical status values, including: For any fault type of electrical equipment, when the data sampling period arrives, the historical state values ​​corresponding to all influencing factors under the fault type are collected to obtain the historical state value corresponding to each fault type; The historical state values ​​corresponding to each fault type are normalized to obtain the historical state values ​​after preprocessing.

3. The electrical management method based on artificial intelligence according to claim 1 or 2, characterized in that: All influencing factors include temperature change rate, ambient temperature, voltage, voltage overvoltage rate, current, current overcurrent rate, internal resistance of electrical equipment, and internal resistance change rate of electrical equipment.

4. The electrical management method based on artificial intelligence according to claim 1, characterized in that: Discretize the historical state values ​​after preprocessing to obtain the discretized historical state values, and perform association support analysis on the discretized historical state values ​​to determine the association support analysis results, including: For the historical state values ​​after preprocessing, the influencing factors are paired in pairs to obtain multiple influencing factor pairing combinations, and the historical state values ​​corresponding to the influencing factor pairing combinations are combined into coordinate points to obtain the discretized historical state values; For the discretized historical state values ​​in the discrete space, a K-means clustering algorithm is used to perform clustering processing to determine the clustering results; wherein, the number of K-means clustering results is the same as the number of fault types; For any fault type, based on the clustering result, the correlation support of the influencing factor pairing combination to the fault type is determined to obtain the correlation support analysis result.

5. The electrical management method based on artificial intelligence according to claim 4 is characterized in that: For the historical state values ​​after preprocessing, the influencing factors are paired in pairs to obtain multiple influencing factor pairing combinations, and the historical state values ​​corresponding to the influencing factor pairing combinations are combined into coordinate points to obtain the discretized historical state values, including: Put all influencing factors into a set, randomly select an influencing factor, and pair the selected influencing factor with other influencing factors in the set to obtain a pairing combination of influencing factors; Determine whether the remaining influencing factors in the set are zero. If so, complete the pairing of the influencing factors to obtain the influencing factor pairing combination corresponding to each extracted influencing factor. Otherwise, return to the step of obtaining the influencing factor pairing combination. For any pairing combination of influencing factors, the historical state value at any historical moment corresponding to the pairing combination of influencing factors is used as the abscissa point and the ordinate point respectively to obtain the discretized historical state value.

6. The electrical management method based on artificial intelligence according to claim 5, characterized in that: For the discretized historical state values ​​in the discrete space, the K-means clustering algorithm is used to perform clustering processing to determine the clustering results, including: Based on the number of fault types K, K cluster centers are randomly selected; where K is a positive integer; For the discretized historical state values ​​in the discrete space, the Euclidean distance between each coordinate point and the cluster center is obtained, and the coordinate point is classified into the cluster center according to the Euclidean distance; When all coordinate points are classified, calculate the average values ​​of all data in K classes respectively, obtain K average values ​​and use them as new cluster centers; Determine whether the cluster center does not change for multiple consecutive times. If so, output the clustering result, otherwise return to the step of classifying the coordinate point into the cluster center.

7. The electrical management method based on artificial intelligence according to claim 6, characterized in that: For any fault type, based on the clustering result, the correlation support of the influencing factor pairing combination to the fault type is determined to obtain the correlation support analysis result, including: For any fault type, based on the clustering result, determine the number of coordinate points related to the fault type in each cluster category; According to the number of coordinate points about the fault type in each cluster category, determine the fault type with the largest number of coordinate points as the real fault type corresponding to the cluster category; For any fault type, determine the cluster category with the same real fault type and the largest number of coordinate points as the target cluster category; For the target clustering category, the proportion of the number of coordinate points of the real fault type in the target clustering category is determined, and the result of the association support analysis is obtained.

8. The electrical management method based on artificial intelligence according to claim 7, characterized in that: Based on the association support analysis results, the target influencing factors corresponding to each fault type are determined, and artificial intelligence is used to learn the historical state values ​​corresponding to the target influencing factors to obtain fault identification rules, including: Determine whether the correlation support analysis result is greater than a preset support threshold, if so, take the influencing factor corresponding to the coordinate point as the target influencing factor corresponding to the real fault type, otherwise do not take the influencing factor corresponding to the coordinate point as the target influencing factor corresponding to the real fault type; Obtain the target influencing factor corresponding to each fault type, and use the historical state value of the target influencing factor at any historical moment as training data, and the corresponding fault type as a training label; A neural network is used to initialize the fault identification rules, and the fault identification rules are trained using training data and training labels to obtain the trained fault identification rules.

9. The electrical management method based on artificial intelligence according to claim 8, characterized in that: Collect the real-time status value corresponding to the target influencing factor, and use the fault identification rule to identify the real-time status value, determine the fault diagnosis result of the electrical equipment, and issue an early warning according to the fault diagnosis result, including: Collect the real-time status values ​​corresponding to the target influencing factors, and form the real-time status values ​​corresponding to the target influencing factors into fault identification rule inputs, and identify the faults through the fault identification rules to obtain fault diagnosis results; When the fault diagnosis result is an abnormal category, a text warning message is generated and transmitted to a device designated by the staff to implement a warning.

10. An electrical management system based on artificial intelligence, characterized in that: include: Data acquisition module, data analysis module, artificial intelligence model and electrical management module; The data acquisition module is used to collect historical status values ​​corresponding to all influencing factors under any fault type of the electrical equipment, and preprocess the historical status values ​​to obtain the historical status values ​​after preprocessing; The data analysis module is used to discretize the historical state values ​​after preprocessing to obtain the discretized historical state values, and to perform an association support analysis on the discretized historical state values ​​to determine the association support analysis results; The artificial intelligence model is used to determine the target influencing factors corresponding to each fault type based on the association support analysis results, and use artificial intelligence to learn the historical state values ​​corresponding to the target influencing factors to obtain fault identification rules; The electrical management module is used to collect real-time status values ​​corresponding to target influencing factors, and use fault identification rules to identify the real-time status values, determine fault diagnosis results of electrical equipment, and issue early warnings based on the fault diagnosis results to complete electrical management.

Images