Fault prediction method and device for power equipment, terminal equipment and storage medium

Abstract

The invention discloses a power equipment fault prediction method and device, terminal equipment and a storage medium, and the method comprises the steps: converting the operation state information of a plurality of pieces of power equipment at a plurality of moments and the external environment information into environment time sequence data and operation time sequence data, and carrying out the fault prediction of the power equipment according to the environment time sequence data and the operation time sequence data; the method comprises the steps of calculating correlation and lagging correlation according to the geographic position of each power device, then dividing a plurality of monitoring areas according to the geographic position of each power device, generating an area device data set of each monitoring area, and finally inputting the area device data set into a preset fault prediction model to generate a fault prediction result of each power device in each monitoring area. According to the method, the influence of external environmental factors on the faults of the power equipment besides the influence of the running state of the power equipment is considered, and the mutual influence of the power equipment in the same range in the running process is also considered, so that monitoring and fault prediction are effectively performed on the power equipment.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a fault prediction method, device, terminal equipment and storage medium for electric power equipment. Background Art

[0002] The power system is the basic guarantee for national economic development and people's livelihood, and its safe and stable operation is of vital importance. The power system is composed of many power generation, transmission, substation, distribution and other equipment assets. The health of these power equipment directly affects the reliability of the entire power grid. Traditional power equipment management mainly relies on manual inspections and regular maintenance, which has problems such as low efficiency and incomplete coverage. With the continuous expansion of the scale of the power grid and the increasing complexity of its structure, this manual management method has been unable to meet the needs.

[0003] Challenges faced by equipment management and safety monitoring of power systems. Various power equipment are widely distributed in power systems, including power plants, substations, transmission lines, etc. The status and operation of these power equipment are crucial to the safety and stability of the entire power grid. However, due to the large number of power equipment and their dispersed geographical locations, traditional manual inspection and monitoring methods can no longer meet the requirements. How to monitor power equipment and predict the risk of power equipment failure has become a key issue that needs to be solved urgently. Summary of the invention

[0004] The embodiments of the present invention provide a method, an apparatus, a terminal device and a storage medium for predicting faults of electric power equipment, which can effectively predict faults of electric power equipment.

[0005] An embodiment of the present invention provides a method for predicting a fault of an electric power device, comprising:

[0006] Obtaining the geographical locations of several power equipment, the operating status information of several power equipment at several times, and the external environment information;

[0007] Generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information;

[0008] According to the environmental time series data and the operating time series data, the correlation and the lag correlation between the environmental time series data and the operating time series data of each power device are calculated; wherein the correlation is the correlation between the environmental time series data and the operating time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operating time series data at different time;

[0009] According to the geographical location, a plurality of monitoring areas are divided; wherein each of the monitoring areas includes a plurality of power equipment;

[0010] Generate a regional equipment data collection of each monitoring area according to the environmental time series data, the operating time series data, the correlation, and the lag correlation of each power equipment in each monitoring area;

[0011] The regional equipment data collection is input into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model for predicting the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

[0012] Furthermore, the geographical location is divided into several monitoring areas, including:

[0013] According to the geographical location, the distance between each of the power equipment is calculated, and a number of initial monitoring areas are divided according to the distance; wherein each initial monitoring area includes a number of power equipment;

[0014] According to the geographical locations of several power equipment in each initial monitoring area, the spatial correlation between the initial monitoring areas is calculated;

[0015] According to the spatial correlation, the plurality of initial monitoring areas are re-divided into a plurality of monitoring areas.

[0016] Further, generating environmental time series data and operating time series data of each power device according to the operating status information and the external environment information includes:

[0017] The operating status information generates the operating sequence data of each power device according to the following formula:

[0018] X k (t) = c k +φ k,1 X k (t-1)+φ k,2 X k (t-2)+…+φ k,p X k (tp)+

[0019] θ k,1 e k (t-1)+θ k,2 e k (t-2)+…+θ k,q e k (tq)+e k (t), k∈(1,2), i∈

[0020] (1,2,…,p), j∈(1,2,…,q);

[0021] Among them, X1(t) is the operation time series data, X2(t) is the environment time series data, φ k,i is the autoregressive coefficient, θ k,j is the moving average coefficient, e k (t) is the white noise at time t, c1 is the preset reference value of the running time series data, and c2 is the preset reference value of the environmental time series data.

[0022] Furthermore, the calculating of the spatial correlation between the initial monitoring areas according to the geographical locations of the plurality of power equipment in the initial monitoring areas includes:

[0023] Determine the target power equipment in each initial monitoring area according to the geographical locations of several power equipment in each initial monitoring area; wherein the target power equipment is the power equipment closest to the center of the initial monitoring area where the target power equipment is located;

[0024] According to the geographical locations of the plurality of electric power equipment in each initial monitoring area, calculating the average geographical locations of the plurality of electric power equipment in each monitoring area;

[0025] According to the geographical location and the average geographical location of the target electric power equipment in each of the initial monitoring areas, the spatial correlation between the initial monitoring areas is calculated.

[0026] Further, the calculating of the spatial correlation between the initial monitoring areas according to the geographical location and the average geographical location of the target power equipment in each initial monitoring area includes:

[0027] The spatial correlation between the initial monitoring areas is calculated according to the following formula:

[0028]

[0029] Among them, r is the spatial correlation coefficient, x i is the geographical location of the target power equipment in the i-th monitoring area, y i is the geographical location of the target power equipment in the jth monitoring area, is the average geographical location of the ith monitoring area, is the average geographical location of the jth monitoring area; w ij is the spatial weight between the ith monitoring area and the jth monitoring area, and n is the total number of monitoring areas.

[0030] Furthermore, the calculation of the correlation and the lag correlation between the environmental time series data and the operation time series data of each power device includes:

[0031] According to the following formula, the correlation between the environmental time series data and the operating time series data of each power device is calculated:

[0032]

[0033] Where R is the correlation, X is the environmental time series data, and X mean is the first average value of the environmental time series data of several power equipment in the monitoring area where the power equipment is located, Y is the operating time series data, and Y mean is a second average value of the operation sequence data of a plurality of electric power equipments in the monitoring area where the electric power equipments are located;

[0034] According to the following formula, the lag correlation between the environmental time series data and the operating time series data of each power equipment is calculated:

[0035]

[0036] Among them, CFF(τ) is the lagged correlation between the environmental time series data and the operating time series data of the power equipment at a lag of τ moments, X(t) is the environmental time series data, and Y(t+τ) is the operating time series data at a lag of τ moments.

[0037] Further, the fault prediction model includes: a first prediction sub-model trained based on a support vector machine, and a second prediction sub-model trained based on a random forest tree;

[0038] The step of inputting the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area includes:

[0039] Inputting the regional equipment data collection into the first prediction sub-model, so that the first prediction sub-model outputs a first failure probability of each power equipment in each monitoring area;

[0040] Inputting the regional equipment data collection into the second prediction sub-model, so that the second prediction sub-model outputs a second failure probability of each power equipment in each monitoring area;

[0041] Performing weighted fusion on the first fault probability and the second fault probability to generate a target fault probability for each power device;

[0042] When the target failure probability is greater than a preset failure probability threshold, determining that the failure prediction result of the electric power equipment is that there is a failure risk;

[0043] When the target failure probability is not greater than a preset failure probability threshold, it is determined that the failure prediction result of the electric power equipment is that there is no failure risk.

[0044] Another embodiment of the present invention provides a fault prediction device for electric power equipment, comprising:

[0045] A data acquisition module, used to acquire the geographical locations of several electric power equipment, the operating status information of several electric power equipment at several times, and external environment information;

[0046] A data processing module, used to generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information;

[0047] A data calculation module, used to calculate the correlation and lag correlation between the environmental time series data and the operation time series data of each power device according to the environmental time series data and the operation time series data; wherein the correlation is the correlation between the environmental time series data and the operation time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operation time series data at different time;

[0048] A region division module, used to divide a number of monitoring regions according to the geographical location; wherein each of the monitoring regions includes a number of power equipment;

[0049] A data integration module, for generating a regional equipment data collection of each monitoring area according to the environmental time series data, the operating time series data, the correlation, and the lag correlation of each power equipment in each monitoring area;

[0050] A fault prediction module is used to input the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model used to predict the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

[0051] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a fault prediction method for an electric power equipment as described in any one of the embodiments is implemented.

[0052] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a fault prediction method for power equipment as described in any one of the above embodiments.

[0053] The following beneficial effects are achieved by implementing the present invention:

[0054] The present invention discloses a method, device, terminal device and storage medium for predicting the fault of an electric device. The method converts the operating status information of several electric devices at several moments and the external environment information into environmental time series data and operating time series data, and calculates the correlation and lag correlation between the environmental time series data and the operating time series data according to the environmental time series data and the operating time series data, and then classifies each electric device into several monitoring areas according to the geographical location of each electric device, and generates a regional device data collection of each monitoring area, and finally inputs the regional device data collection into a preset fault prediction model to generate a fault prediction result of each electric device in each monitoring area. The present invention not only considers that the fault of the electric device is affected by external environmental factors in addition to its own operating state, but also considers that the electric devices in the same range will affect each other during operation. Therefore, the present invention calculates the correlation and lag correlation between the environmental time series data and the operating time series data, and classifies the electric device into several monitoring areas to perform subsequent fault prediction, and finally adopts the fault prediction model to predict the fault risk of each electric device in each monitoring area, so as to effectively monitor and predict the fault of the electric device. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of a method for predicting faults of electric power equipment provided in one embodiment of the present invention.

[0056] Figure 2 It is a structural schematic diagram of a fault prediction device for electric power equipment provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0059] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0060] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0061] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0062] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0063] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the internal connection of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0064] See also Figure 1 , is a flow chart of a method for predicting a fault of an electric power device provided by an embodiment of the present invention, comprising:

[0065] S1. Obtaining the geographical locations of several power equipment, the operating status information of several power equipment at several times, and the external environment information;

[0066] In a preferred embodiment of the present invention, the power equipment includes transmission lines, transformers, converters, etc., and the operating status information of each power equipment at certain moments and external environment information are collected based on preset networkable sensors, and then the electronic equipment is extracted through active detection and DPI detection technology based on application identification and protocol identification capabilities; through network traffic analysis, the specific application running in the sensor device is identified. This can be achieved by detecting specific application layer protocols, port numbers, data content, etc. in network data packets.

[0067] Or, by analyzing the network protocol used by the IoT device, determine the type of communication protocol used by the device (such as HTTP, MQTT, CoAP, etc.). Protocol identification usually relies on deep packet inspection (DPI) technology, or, by parsing the protocol header and payload of the network data packet to identify the protocol type. Use active detection technology (such as ICMP detection, port scanning, SNMP query, etc.) to identify and collect information about IoT devices in the network. Or, active detection can actively send requests to the device and analyze the response to obtain the device's IP address, MAC address, operating system type, open ports and other information. Or, by scanning all online devices in the LAN or WAN, determine the existence of IoT devices and extract their basic information. This process can be achieved using tools such as Nmap or specially customized scanners. Or, DPI technology extracts application layer data by deeply analyzing the content of network data packets. DPI can parse the application layer protocol, identify key fields in the data, and extract information such as device type, protocol version, application status, etc. Or, combining traffic pattern analysis and DPI technology, network traffic can be continuously monitored and analyzed to identify abnormal behavior or potential security threats. At the same time, the application status information of IoT devices can be extracted in real time.

[0068] Further, based on the information extracted in step 1, the electronic devices are identified and classified. Basic information includes device ID, type, manufacturer, firmware version, network interface information, geographic location, etc. Based on the identification and classification of the device, its metadata information is further extracted, such as operating time, usage status, firmware update records, configuration files, etc. This information can be obtained through the device management interface or SNMP and other protocols. Environmental data is collected from environmental sensors deployed around the IoT devices. These data may include temperature, humidity, air pressure, light, vibration, etc. These sensors usually communicate with the devices through sensor networks or IoT platforms. The data collected by the sensors is integrated with the basic information of the power equipment to form a comprehensive data set of the equipment and the environment. This process needs to consider data synchronization and timing alignment to ensure data consistency. The collected basic information of the power equipment and environmental data are pre-processed by cleaning, format conversion, data verification and other pre-processing steps to ensure data accuracy and consistency. The pre-processed data is uploaded to the asset management platform through a secure communication channel (such as HTTPS, MQTT, etc.) to prepare for subsequent data fusion and analysis.

[0069] Furthermore, the wireless gateway device extracts IoT asset information through active detection and DPI detection technology. Active detection technology is used to identify and collect IoT device information in the network, including basic information such as IP address, MAC address, operating system type, open ports, etc.; DPI technology deeply analyzes network traffic data packets and extracts application layer data of devices, such as device type, protocol version, application status, etc. The collected asset information will be pre-processed by the wireless gateway device, including format conversion, data verification and outlier detection to ensure data quality and consistency; including format conversion, data verification, outlier detection and other operations on the collected data to ensure data quality. The pre-processed asset information and environmental data are integrated to form a complete asset status information set.

[0070] The asset status information after preprocessing and fusion is first temporarily stored in the local cache area of ​​the wireless gateway device. For the data in the cache area, a reasonable storage time and capacity threshold are set. When the time or capacity is exceeded, the data is cleaned and uploaded.

[0071] Regularly upload the locally cached asset status information to the centralized power asset management platform through encrypted transmission. Ensure the security of the data transmission process, such as using SSL / TLS encryption, digital signatures, etc. For abnormal situations that occur during the transmission process, such as network interruptions, implement a temporary storage and retransmission mechanism for local data.

[0072] The power asset management platform receives asset status information from the wireless gateway, stores and manages it, and establishes a full life cycle information archive of power assets, including location information, operating status, maintenance records, etc., providing reliable basic data support for subsequent data fusion and intelligent analysis.

[0073] Through the above steps, the wireless gateway device can pre-process and temporarily store the collected basic asset information and environmental data, and upload them to the power asset management platform in a safe and reliable manner. This process ensures data quality and transmission security, laying the foundation for subsequent data fusion and analysis.

[0074] Furthermore, the power asset management platform needs to connect and exchange data with other information systems such as the power grid dispatching system and the security monitoring system. In view of the differences in data formats and transmission protocols of different systems, a unified data access mechanism is adopted, such as a standard API-based method, to achieve cross-system data exchange. All types of received data are standardized in a unified data format to lay the foundation for subsequent data fusion.

[0075] The asset management platform associates and maps the standardized data with the local asset ledger based on key identification information such as asset number and equipment location. It establishes an association model between power assets, grid operation status, and security monitoring data, and associates and merges these originally independent data. It forms a comprehensive data set that includes comprehensive information such as basic asset information, environmental data, grid status, and security events.

[0076] S2. Generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information;

[0077] Preferably, generating environmental time series data and operating time series data of each electric power device according to the operating state information and the external environment information includes:

[0078] S21. Generate the operation sequence data of each power device based on the operation status information according to the following formula:

[0079] X k (t) = c k +φ k,1 X k (t-1)+φ k,2 X k (t-2)+…+φ k,p X k (tp)+

[0080] θ k,1 e k (t-1)+θ k,2 ek (t-2)+…+θ k,q e k (tq)+e k (t), k∈(1,2), i∈

[0081] (1,2,…,p), j∈(1,2,…,q);

[0082] Among them, X1(t) is the operation time series data, X2(t) is the environment time series data, φ k,i is the autoregressive coefficient, θ k,j is the moving average coefficient, e k (t) is the white noise at time t, c1 is the preset reference value of the running time series data, and c2 is the preset reference value of the environmental time series data.

[0083] In a preferred embodiment of the present invention, a time series analysis method is used to generate environmental time series data and operating time series data of each power device. Specifically,

[0084] X k (t) = c k +φ k,1 X k (t-1)+φ k,2 X k (t-2)+…+φ k,p X k (tp)+

[0085] θ k,1 e k (t-1)+θ k,2 e k (t-2)+…+θ k,q e k (tq)+e k (t), k∈(1,2), i∈

[0086] (1,2,…,p), j∈(1,2,…,q);

[0087] Among them, X1(t) is the operation time series data, X2(t) is the environment time series data, φ k,i is the autoregressive coefficient, θ k,j is the moving average coefficient, e k (t) is the white noise at time t, c1 is the preset reference value of the running time series data, and c2 is the preset reference value of the environmental time series data.

[0088] S3. Calculate the correlation and lag correlation between the environmental time series data and the operation time series data of each power device according to the environmental time series data and the operation time series data; wherein the correlation is the correlation between the environmental time series data and the operation time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operation time series data at different time;

[0089] Preferably, the calculation of the correlation and hysteresis correlation between the environmental time series data and the operation time series data of each power device includes:

[0090] S31. Calculate the correlation between the environmental time series data and the operating time series data of each power device according to the following formula:

[0091]

[0092] Where R is the correlation, X is the environmental time series data, and X mean is the first average value of the environmental time series data of several power equipment in the monitoring area where the power equipment is located, Y is the operating time series data, and Y mean is a second average value of the operation sequence data of a plurality of electric power equipments in the monitoring area where the electric power equipments are located;

[0093] S32. Calculate the lag correlation between the environmental time series data and the operating time series data of each power device according to the following formula:

[0094]

[0095] Among them, CFF(τ) is the lagged correlation between the environmental time series data and the operating time series data of the power equipment at a lag of τ moments, X(t) is the environmental time series data, and Y(t+τ) is the operating time series data at a lag of τ moments.

[0096] In a preferred embodiment of the present invention, based on the environmental time series data and the operation time series data generated in step S2, the correlation between the environmental time series data and the operation time series data is mined. Specifically, the correlation between the environmental time series data and the operation time series data of each power device is calculated according to the following formula:

[0097]

[0098] Where R is the correlation, X is the environmental time series data, and X mean is the first average value of the environmental time series data of several power equipment in the monitoring area where the power equipment is located, Y is the operating time series data, and Y mean is a second average value of the operation sequence data of a plurality of electric power equipments in the monitoring area where the electric power equipments are located;

[0099] Secondly, the cross-correlation function is used to calculate the time correlation between different data indicators to reflect the correlation degree of the two time series at different time lags. That is, the lag correlation between the environmental time series data and the operation time series data of each power equipment is calculated according to the following formula:

[0100]

[0101] Among them, CFF(τ) is the lagged correlation between the environmental time series data and the operating time series data of the power equipment at a lag of τ moments, X(t) is the environmental time series data, and Y(t+τ) is the operating time series data at a lag of τ moments.

[0102] It should be noted that τ represents the size of the time lag and can be positive or negative.

[0103] S4. Divide a plurality of monitoring areas according to the geographical location; wherein each of the monitoring areas includes a plurality of power equipment;

[0104] Preferably, the plurality of monitoring areas are divided according to the geographical location, including:

[0105] S41, calculating the distance between each of the power devices according to the geographical location, and dividing a number of initial monitoring areas according to the distance; wherein each initial monitoring area includes a number of power devices;

[0106] S42, calculating the spatial correlation between the initial monitoring areas according to the geographical locations of a number of power equipment in each initial monitoring area;

[0107] Preferably, the calculating of the spatial correlation between the initial monitoring areas according to the geographical locations of a number of power equipment in each initial monitoring area includes:

[0108] S421, determining the target power equipment in each initial monitoring area according to the geographical locations of several power equipment in each initial monitoring area; wherein the target power equipment is a power equipment closest to the center of the initial monitoring area where the target power equipment is located;

[0109] S422, calculating the average geographical location of the plurality of electric power equipment in each monitoring area according to the geographical location of the plurality of electric power equipment in each initial monitoring area;

[0110] S423: Calculate the spatial correlation between the initial monitoring areas according to the geographical location and the average geographical location of the target power equipment in each of the initial monitoring areas.

[0111] Preferably, the spatial correlation between the initial monitoring areas is calculated according to the following formula:

[0112]

[0113] Among them, r is the spatial correlation coefficient, x i is the geographical location of the target power equipment in the i-th monitoring area, y i is the geographical location of the target power equipment in the jth monitoring area, is the average geographical location of the ith monitoring area, is the average geographical location of the jth monitoring area; w ij is the spatial weight between the ith monitoring area and the jth monitoring area, and n is the total number of monitoring areas.

[0114] S43. Re-divide the initial monitoring areas into a plurality of monitoring areas according to the spatial correlation.

[0115] In a preferred embodiment of the present invention, the geographical locations of the power equipment are mapped into a geographic space model to determine their distribution, and then the distances between the power equipment are calculated, and a number of initial monitoring areas are divided according to the distances;

[0116] Secondly, calculate the spatial correlation coefficient of each initial monitoring area:

[0117]

[0118] Among them, r is the spatial correlation coefficient, x i is the geographical location of the target power equipment in the i-th monitoring area, y i is the geographical location of the target power equipment in the jth monitoring area, is the average geographical location of the ith monitoring area, is the average geographical location of the jth monitoring area; w ij is the spatial weight between the ith monitoring area and the jth monitoring area, and n is the total number of monitoring areas.

[0119] The spatial correlation coefficient r is used as the input feature of the clustering algorithm, and the clustering algorithm is used to merge the areas with higher correlation into one monitoring area:

[0120]

[0121] Where: C i represents the i-th cluster; μ i represents the i-th cluster center; r represents the spatial correlation coefficient; by minimizing the objective function J, k monitoring areas are obtained.

[0122] S5. Generate a regional equipment data collection of each monitoring area according to the environmental time series data, the operation time series data, the correlation, and the lag correlation of each power equipment in each monitoring area;

[0123] S6. Input the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model for predicting the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

[0124] Preferably, the fault prediction model includes: a first prediction sub-model trained based on a support vector machine, and a second prediction sub-model trained based on a random forest tree;

[0125] The step of inputting the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area includes:

[0126] S61, inputting the regional equipment data collection into the first prediction sub-model, so that the first prediction sub-model outputs a first failure probability of each power equipment in each monitoring area;

[0127] S62, inputting the regional equipment data collection into the second prediction sub-model, so that the second prediction sub-model outputs a second failure probability of each power equipment in each monitoring area;

[0128] S63, performing weighted fusion on the first fault probability and the second fault probability to generate a target fault probability for each power device;

[0129] S64, when the target failure probability is greater than a preset failure probability threshold, determining that the failure prediction result of the power equipment is that there is a failure risk;

[0130] S65. When the target failure probability is not greater than a preset failure probability threshold, determine that the failure prediction result of the power equipment is that there is no failure risk.

[0131] In a preferred embodiment of the present invention, during training, the regional equipment data set is used as the input variable of the support vector machine model, and the failure risk of the power equipment is used as the output variable; and the support vector machine is trained according to a preset objective function;

[0132] Among them, the objective function of the support vector machine is as follows:

[0133]

[0134] Where: w is the model parameter, b is the bias term; ε i is a slack variable; C is a penalty factor; by optimizing the objective function, the optimal hyperplane parameters w and b are obtained for classification prediction of new input data;

[0135] Random forest model: The input features are used as input variables of the random forest model, and the failure risk of power equipment is used as the output variable; the random forest model includes multiple decision trees, each of which is trained based on a randomly selected feature subset; the random forest model finally outputs the majority vote or average of the outputs of all decision trees to obtain the final classification prediction result;

[0136] The prediction results of the support vector model and the random forest model are fused by weighted averaging to obtain the final fault risk prediction result.

[0137] The present embodiment provides a method for predicting the fault of an electric device, by converting the operating status information of several electric devices at several moments and the external environment information into environmental time series data and operating time series data, and calculating the correlation and lag correlation between the environmental time series data and the operating time series data according to the environmental time series data and the operating time series data, and then classifying each electric device into several monitoring areas according to the geographical location of each electric device, and generating a regional equipment data collection of each monitoring area, and finally inputting the regional equipment data collection into a preset fault prediction model to generate a fault prediction result for each electric device in each monitoring area. The present embodiment not only considers that the fault of the electric device is affected by external environmental factors in addition to its own operating state, but also considers that the electric devices in the same range will affect each other during operation. Therefore, the present embodiment calculates the correlation and lag correlation between the environmental time series data and the operating time series data, and classifies the electric device into several monitoring areas to perform subsequent fault prediction, and finally adopts the fault prediction model to predict the fault risk of each electric device in each monitoring area, so as to effectively monitor and predict the fault of the electric device.

[0138] See also Figure 2 , is a schematic diagram of a structure of a fault prediction device for electric power equipment provided by an embodiment of the present invention, comprising:

[0139] A data acquisition module, used to acquire the geographical locations of several electric power equipment, the operating status information of several electric power equipment at several times, and external environment information;

[0140] A data processing module, used to generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information;

[0141] A data calculation module, used to calculate the correlation and lag correlation between the environmental time series data and the operation time series data of each power device according to the environmental time series data and the operation time series data; wherein the correlation is the correlation between the environmental time series data and the operation time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operation time series data at different time;

[0142] A region division module, used to divide a number of monitoring regions according to the geographical location; wherein each of the monitoring regions includes a number of power equipment;

[0143] A data integration module, for generating a regional equipment data collection of each monitoring area according to the environmental time series data, the operating time series data, the correlation, and the lag correlation of each power equipment in each monitoring area;

[0144] A fault prediction module is used to input the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model used to predict the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

[0145] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0146] Those skilled in the art can clearly understand that for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0147] Another preferred embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements a fault prediction method for an electric power equipment as described in any one of the above embodiments.

[0148] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0149] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0150] The memory can be used to store the computer program, and the processor realizes various functions of the terminal device by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0151] Another preferred embodiment of the present invention provides a storage medium, the storage medium is a computer-readable storage medium, the computer program is stored in the computer-readable storage medium, and the computer program, when executed by the processor, can implement the steps of the above-mentioned various method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0152] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for predicting a fault of an electric power device, characterized in that: include: Obtaining the geographical locations of several power equipment, the operating status information of several power equipment at several times, and the external environment information; Generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information; According to the environmental time series data and the operating time series data, the correlation and the lag correlation between the environmental time series data and the operating time series data of each power device are calculated; wherein the correlation is the correlation between the environmental time series data and the operating time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operating time series data at different time; According to the geographical location, a plurality of monitoring areas are divided; wherein each of the monitoring areas includes a plurality of power equipment; Generate a regional equipment data collection of each monitoring area according to the environmental time series data, the operating time series data, the correlation, and the lag correlation of each power equipment in each monitoring area; The regional equipment data collection is input into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model for predicting the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

2. A method for predicting faults of electric power equipment according to claim 1, characterized in that: According to the geographical location, several monitoring areas are divided, including: According to the geographical location, the distance between each of the power equipment is calculated, and a number of initial monitoring areas are divided according to the distance; wherein each initial monitoring area includes a number of power equipment; According to the geographical locations of several power equipment in each initial monitoring area, the spatial correlation between the initial monitoring areas is calculated; According to the spatial correlation, the plurality of initial monitoring areas are re-divided into a plurality of monitoring areas.

3. A method for predicting faults of electric power equipment according to claim 2, characterized in that: The step of generating environmental time series data and operating time series data of each electric power device according to the operating state information and the external environment information includes: The operating status information generates the operating sequence data of each power device according to the following formula: Among them, X1(t) is the operation time series data, X2(t) is the environment time series data, φ k,i is the autoregressive coefficient, θ k,j is the moving average coefficient, e k (t) is the white noise at time t, c1 is the preset reference value of the running time series data, and c2 is the preset reference value of the environmental time series data.

4. A method for predicting faults of electric power equipment according to claim 3, characterized in that: The calculating of the spatial correlation between the initial monitoring areas according to the geographical locations of the plurality of power equipment in the initial monitoring areas comprises: Determine the target power equipment in each initial monitoring area according to the geographical locations of several power equipment in each initial monitoring area; wherein the target power equipment is the power equipment closest to the center of the initial monitoring area where the target power equipment is located; According to the geographical locations of the plurality of electric power equipment in each initial monitoring area, calculating the average geographical locations of the plurality of electric power equipment in each monitoring area; According to the geographical location and the average geographical location of the target electric power equipment in each of the initial monitoring areas, the spatial correlation between the initial monitoring areas is calculated.

5. A method for predicting faults of electric power equipment according to claim 4, characterized in that: The calculating the spatial correlation between the initial monitoring areas according to the geographical location and the average geographical location of the target power equipment in each initial monitoring area comprises: The spatial correlation between the initial monitoring areas is calculated according to the following formula: Among them, r is the spatial correlation coefficient, x i is the geographical location of the target power equipment in the i-th monitoring area, y i is the geographical location of the target power equipment in the jth monitoring area, is the average geographical location of the ith monitoring area, is the average geographical location of the jth monitoring area; w ij is the spatial weight between the ith monitoring area and the jth monitoring area, and n is the total number of monitoring areas.

6. A method for predicting faults of electric power equipment according to claim 5, characterized in that: The calculating of the correlation and the lag correlation between the environmental time series data and the operation time series data of each electric power device includes: According to the following formula, the correlation between the environmental time series data and the operating time series data of each power device is calculated: Where R is the correlation, X is the environmental time series data, and X mean is the first average value of the environmental time series data of several power equipment in the monitoring area where the power equipment is located, Y is the operating time series data, and Y mean is a second average value of the operation sequence data of a plurality of electric power equipments in the monitoring area where the electric power equipments are located; According to the following formula, the lag correlation between the environmental time series data and the operating time series data of each power equipment is calculated: Among them, CFF(τ) is the lagged correlation between the environmental time series data and the operating time series data of the power equipment at a lag of τ moments, X(t) is the environmental time series data, and Y(t+τ) is the operating time series data at a lag of τ moments.

7. A method for predicting faults of electric power equipment according to claim 6, characterized in that: The fault prediction model includes: a first prediction sub-model trained based on a support vector machine, and a second prediction sub-model trained based on a random forest tree; The step of inputting the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area includes: Inputting the regional equipment data collection into the first prediction sub-model, so that the first prediction sub-model outputs a first failure probability of each power equipment in each monitoring area; Inputting the regional equipment data collection into the second prediction sub-model, so that the second prediction sub-model outputs a second failure probability of each power equipment in each monitoring area; Performing weighted fusion on the first fault probability and the second fault probability to generate a target fault probability for each power device; When the target failure probability is greater than a preset failure probability threshold, determining that the failure prediction result of the electric power equipment is that there is a failure risk; When the target failure probability is not greater than a preset failure probability threshold, it is determined that the failure prediction result of the electric power equipment is that there is no failure risk.

8. A fault prediction device for electric power equipment, characterized in that: include: A data acquisition module, used to acquire the geographical locations of several electric power equipment, the operating status information of several electric power equipment at several times, and external environment information; A data processing module, used to generate environmental time series data and operating time series data of each power device according to the operating status information and the external environment information; A data calculation module, used to calculate the correlation and lag correlation between the environmental time series data and the operation time series data of each power device according to the environmental time series data and the operation time series data; wherein the correlation is the correlation between the environmental time series data and the operation time series data at the corresponding time; the lag correlation is the correlation between the environmental time series data and the operation time series data at different time; A region division module, used to divide a number of monitoring regions according to the geographical location; wherein each of the monitoring regions includes a number of power equipment; A data integration module, for generating a regional equipment data collection of each monitoring area according to the environmental time series data, the operating time series data, the correlation, and the lag correlation of each power equipment in each monitoring area; A fault prediction module is used to input the regional equipment data collection into a preset fault prediction model so that the fault prediction model outputs a fault prediction result for each power equipment in each monitoring area; wherein the fault prediction result is used to characterize whether there is a failure risk of the power equipment; the fault prediction model is a pre-trained machine learning model used to predict the failure risk of each power equipment in the monitoring area based on the regional equipment data collection.

9. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a fault prediction method for an electric power device as claimed in any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is executed, the device where the storage medium is located is controlled to execute a fault prediction method for electric power equipment as described in any one of claims 1 to 7.

Images