Electric power supply fault monitoring system based on artificial intelligence
By applying an artificial intelligence-based fault monitoring system in the power system and using the convolutional neural network model to analyze power parameters, the accurate assessment and automatic diagnosis of fault risks are achieved, and the problem of difficulty in accurately positioning faults in traditional monitoring systems is solved, which improves the problem of the problem of traditional monitoring systems being difficult to accurately locate faults. The accuracy of fault detection and maintenance efficiency of power systems.
Patent Information
- Application Number
- CN202411152561.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-26
AI Technical Summary
Traditional power supply fault monitoring systems lack intelligent analysis methods and are difficult to accurately locate fault locations, which makes troubleshooting time and effort consuming and may miss the best repair time, and cannot meet the requirements of modern power systems for real-time and accuracy.
Design a power supply fault monitoring system based on artificial intelligence, including data acquisition module, data transmission module, data processing module and risk division module. Through the convolutional neural network model, multi-level analysis of power parameters is performed, fault characteristics are extracted, and fault characteristics are constructed. Fault prediction model, evaluates the risk of failure, and automatically issues alarm signals and performs emergency response measures when a fault is detected.
It improves the accuracy and timeliness of fault detection, significantly reduces the maintenance cost of the power system, ensures the stability and reliability of the system, and enhances the fault management capabilities and intelligent operation level.
Smart Images

Figure CN119030142A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power system monitoring, and in particular to an artificial intelligence-based power supply fault monitoring system. Background Art
[0002] With the development of modern society, power supply has become an indispensable part of industrial, commercial and family life. In order to ensure the reliability and stability of the power system, fault monitoring of the power supply system is particularly important. Traditional power supply fault monitoring systems usually rely on manual inspection and traditional monitoring equipment. These methods have many limitations and are difficult to meet the requirements of modern power systems for real-time and accuracy. However, traditional monitoring systems lack intelligent analysis methods and are difficult to accurately locate faults. It is usually necessary to rely on manual experience for troubleshooting, which is not only time-consuming and labor-intensive, but may also miss the best time for maintenance. Summary of the invention
[0003] The purpose of the present invention is to provide an artificial intelligence-based power supply fault monitoring system to address the deficiencies in the background technology.
[0004] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: an artificial intelligence-based power supply fault monitoring system, comprising a data acquisition module, a data transmission module, a data processing module, a risk classification module and a control module;
[0005] A data acquisition module is arranged at several power data acquisition nodes in the power system and is used to collect power parameters of the power system in real time;
[0006] The data transmission module performs multi-level analysis on the collected power parameters through the convolutional neural network model and transmits them to the data processing module;
[0007] The data processing module pre-processes the transmitted power parameters, extracts the fault characteristics in the power parameters and builds a fault prediction model, and evaluates the risk of failure of each power data collection node in the power system;
[0008] A risk classification module, which divides the risk of failure of power data collection nodes in the power system into high-risk collection nodes and low-risk collection nodes according to the evaluation results;
[0009] The control module performs fault diagnosis on high-risk acquisition nodes, automatically sends out alarm signals and executes emergency treatment measures when a fault is detected.
[0010] In a preferred embodiment, in the data transmission module, a multi-level analysis is performed on the collected power parameters through a convolutional neural network model, specifically:
[0011] The real-time collected power parameters are preprocessed and then input into the trained convolutional neural network model;
[0012] The convolutional and pooling layers perform multi-level feature extraction on the input data to identify features and patterns in the data;
[0013] The fully connected layer and the output layer conduct a comprehensive analysis of the extracted features and output the prediction results;
[0014] Determine whether there is a fault in the power system, as well as the type and location of the fault based on the prediction results;
[0015] The output results of the convolutional neural network model are transmitted to the data processing module.
[0016] In a preferred embodiment, the fault characteristics in the power parameters include a power supply drift index of the power equipment and a communication frequency deviation value of the power equipment.
[0017] In a preferred embodiment, the method for obtaining the power supply drift index is:
[0018] Obtain the number j of power supply devices in a time period s, and label them as j = 1, 2, 3...c, where c is a positive integer greater than 0, calculate the power supply of each power supply device, and calculate the power supply efficiency of each power supply device in the power system; establish a set of power supply efficiencies of each power supply device in the time period s, calculate the mean of the set, i.e., the mean of the power supply efficiencies of the power supply devices, analyze the real-time power supply efficiency of each power supply device with the mean of the power supply efficiency of the power supply devices, calculate the standard deviation of the power supply efficiency of all power supply devices in the power system, i.e., calculate the power supply drift index.
[0019] In a preferred embodiment, the method for obtaining the communication frequency deviation value is:
[0020] Set a sliding window w of fixed size, send requests to the power system regularly, record the results, and save the information such as the delay and packet loss rate of each command in the sliding window; perform weighted calculation on each command result in the sliding window to obtain the communication stability value hx of the power equipment, and weight each command result according to its time interval in the sliding window. The delay value is Li, the time interval is Ti, and the weight coefficient is Wi, where i represents the result of the i-th command sending. Calculate the weighted communication delay. The specific calculation expression is: Among them, n is the number of command results in the sliding window, yp is the weighted communication delay, the communication stability value and weighted communication delay of the power equipment are analyzed, and the communication frequency deviation value is calculated. The specific calculation expression is: DF = |hx-yp|; where DF is the communication frequency deviation value.
[0021] In a preferred embodiment, the power supply drift index and the communication frequency deviation value are normalized, and the risk coefficient of failure of each power data acquisition node in the power system is calculated based on the normalized power supply drift index and communication frequency deviation value.
[0022] In a preferred embodiment, in the risk classification module, the risk of failure of the power data collection nodes in the power system is divided into high-risk collection nodes and low-risk collection nodes, specifically:
[0023] The acquired risk coefficient of failure of each power data acquisition node in the power system is compared with the risk coefficient reference threshold value. If the risk coefficient of failure of each power data acquisition node in the power system is greater than or equal to the risk coefficient reference threshold value, the risk of failure of the power data acquisition node in the power system is classified as a high-risk acquisition node, and an early warning signal is generated at this time;
[0024] If the risk coefficient of failure of each power data acquisition node in the power system is less than the risk coefficient reference threshold, the risk of failure of the power data acquisition node in the power system is classified as a low-risk acquisition node, and no warning signal is generated at this time.
[0025] In a preferred embodiment, when the risk of failure of a power data acquisition node in the power system is classified as a high-risk acquisition node, that is, the risk coefficient of failure of each power data acquisition node in the power system within a period of time is greater than or equal to a risk coefficient reference threshold, the risk coefficients greater than or equal to the risk coefficient reference threshold in a subsequent period of time are collected and a corresponding data set is established.
[0026] In a preferred embodiment, for each high-risk node n∈Nhigh, n is the number of risk coefficients in the set, and Nhigh is the risk coefficient set of high-risk collection nodes; its power parameter data Dn is collected in real time and fault diagnosis is performed. If the fault characteristic function F(Dn) is greater than or equal to the fault threshold TF, the node n is judged to be faulty; when a fault is detected in the node n, the system generates an alarm signal An and sends the fault information to relevant personnel, An=alarm signal (n); according to the fault type Fn and the preset emergency handling plan E, the corresponding emergency handling measures are executed, the emergency handling function E(Fn) is defined, and the system executes corresponding measures according to the fault type Fn.
[0027] In the above technical solution, the technical effects and advantages provided by the present invention are:
[0028] 1. The present invention realizes real-time collection of power parameters by arranging multiple power data collection nodes in the power system. The convolutional neural network model is used to perform multi-level analysis on the collected power parameters, extract fault characteristics and build a fault prediction model, which can effectively evaluate the fault risk of each node. The system divides the node risk into two categories: high risk and low risk, performs real-time fault diagnosis on high-risk nodes, and automatically sends out alarm signals and executes emergency treatment measures when a fault is detected. The accuracy and timeliness of fault detection are improved, and the maintenance cost of the power system is significantly reduced.
[0029] 2. The present invention can accurately evaluate the risk of power data acquisition nodes and ensure the stability and reliability of the system by acquiring and normalizing the power supply drift index and communication frequency deviation value. Real-time monitoring and emergency response measures for high-risk nodes effectively prevent the spread of faults and ensure the continuity and safety of power supply. On the whole, it not only enhances the fault management capability of the power system, but also improves the intelligent level of power system operation, with significant economic and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0031] Figure 1 It is a system module diagram of the present invention. DETAILED DESCRIPTION
[0032] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0033] For examples, see Figure 1 As shown, the power supply fault monitoring system based on artificial intelligence in this embodiment includes a data acquisition module, a data transmission module, a data processing module, a risk classification module and a control module;
[0034] A data acquisition module is arranged at several power data acquisition nodes in the power system and is used to collect power parameters of the power system in real time;
[0035] The data transmission module performs multi-level analysis on the collected power parameters through the convolutional neural network model and transmits them to the data processing module;
[0036] The data processing module pre-processes the transmitted power parameters, extracts the fault characteristics in the power parameters and builds a fault prediction model, and evaluates the risk of failure of each power data collection node in the power system;
[0037] A risk classification module, which divides the risk of failure of power data collection nodes in the power system into high-risk collection nodes and low-risk collection nodes according to the evaluation results;
[0038] The control module performs fault diagnosis on high-risk acquisition nodes, automatically sends out alarm signals and executes emergency treatment measures when a fault is detected.
[0039] Among them, in the data acquisition module, several power data acquisition nodes arranged in the power system are used to collect power parameters of the power system in real time;
[0040] Power data acquisition nodes are installed in the substation, and these nodes are arranged on different devices and output ports of the substation. These nodes collect voltage and current parameters of key equipment in the substation to ensure the normal operation of the equipment.
[0041] Power data acquisition nodes are installed in the distribution station to monitor the power parameters in the distribution station. These nodes monitor the current and voltage of each transmission line in the distribution station, as well as environmental parameters such as temperature and humidity.
[0042] On long-distance transmission lines, power data collection nodes are installed at regular intervals. These nodes collect current and voltage data of the transmission lines in real time and monitor the health status of the transmission lines.
[0043] Each data acquisition node is equipped with a variety of sensors to monitor the operating status of the power system in real time. Sensor types include voltage sensors, current sensors, temperature sensors, and humidity sensors.
[0044] The collected data is transmitted to the central data processing system in real time via wireless or wired networks. High-reliability communication technology is used for data transmission to ensure that data is not lost or delayed.
[0045] Voltage sensors are used to monitor voltage changes in power systems to ensure that the voltage is within a safe range. Real-time voltage monitoring can prevent overvoltage or low voltage from damaging equipment.
[0046] Current sensors are used to monitor the flow of current in power systems to ensure that the current is within the normal range. Real-time current monitoring can prevent overcurrent and avoid overloading of power equipment.
[0047] Temperature sensors are used to monitor the operating temperature of power equipment to ensure that the equipment operates within a safe temperature range. Real-time temperature monitoring can prevent equipment failure due to overheating.
[0048] Humidity sensors are used to monitor the humidity of the environment in which power equipment is located to prevent short circuits or corrosion of equipment due to excessive humidity. Real-time humidity monitoring can ensure the suitability of the operating environment of power equipment.
[0049] The data transmission module performs multi-level analysis on the collected power parameters through the convolutional neural network model and transmits them to the data processing module.
[0050] Each power data collection node collects the voltage, current, temperature, humidity and other parameters of the power system in real time. The sensor transmits the collected data to the primary data processing unit through a wireless network or a wired network.
[0051] Clean the collected raw data to remove noise and invalid data. Convert and standardize the data format for subsequent analysis and processing.
[0052] The preprocessed data is divided into windows of fixed size to form time series data blocks. For example, the data collected every minute is divided into a data block, and each data block contains the values of multiple power parameters.
[0053] The data in each data block is normalized to scale the data values to a standard range (such as between 0 and 1).
[0054] Design a convolutional neural network model that includes the following hierarchy: Input layer: receives normalized data blocks.
[0055] Convolution layer: Perform convolution operation on the input data to extract local features.
[0056] Pooling layer: Downsampling is performed to reduce the data dimension and retain important features.
[0057] Stacking of multiple convolutional layers and pooling layers: further extracting higher-level features.
[0058] Fully connected layer: maps the extracted features to the output space.
[0059] Output layer: Outputs prediction results according to specific tasks (such as fault classification and feature extraction).
[0060] The convolutional neural network model is trained using historical power data. A supervised learning method is used, using labeled fault data as the training set. The model parameters are adjusted through the back-propagation algorithm and optimizer (such as Adam) to minimize the loss function.
[0061] Use the validation set to validate the trained model and evaluate the accuracy and robustness of the model.
[0062] Adjust the model structure and parameters according to the verification results to further optimize the model performance.
[0063] The real-time collected power parameters are preprocessed and input into the trained convolutional neural network model.
[0064] The convolutional and pooling layers perform multi-level feature extraction on the input data to identify key features and patterns in the data.
[0065] The fully connected layer and the output layer conduct a comprehensive analysis of the extracted features and output the prediction results.
[0066] Based on the prediction results, determine whether there is a fault in the power system, as well as the type and location of the fault.
[0067] The output results of the convolutional neural network model (including features and prediction results) are transmitted to the data processing module.
[0068] Adopt high-reliability communication technology to ensure the real-time and accuracy of data transmission.
[0069] After receiving the analysis results, the central data processing system stores them in the big data storage system. The historical data and real-time data are managed in a unified manner to facilitate subsequent analysis and query. The data processing system further processes and comprehensively analyzes the received analysis results. A fault report is generated to provide the fault location, type, and possible cause. Decision support suggestions are provided to help operation and maintenance personnel quickly respond to and handle faults.
[0070] The data processing module pre-processes the transmitted power parameters, extracts the fault characteristics in the power parameters and builds a fault prediction model, and evaluates the risk of failure of each power data acquisition node in the power system.
[0071] Among them, the fault characteristics in the power parameters include the power supply drift index of the power equipment and the communication frequency deviation value of the power equipment;
[0072] The method for obtaining the power supply drift index is as follows:
[0073] Obtain the number j of power supply devices in a time period s, and label them as j = 1, 2, 3...c, where c is a positive integer greater than 0, calculate the power supply of each power supply device, and calculate the power supply efficiency of each power supply device in the power system; establish a set of power supply efficiencies of each power supply device in the time period s, calculate the mean of the set, i.e., the mean of the power supply efficiencies of the power supply devices, analyze the real-time power supply efficiency of each power supply device with the mean of the power supply efficiency of the power supply devices, calculate the standard deviation of the power supply efficiency of all power supply devices in the power system, i.e., calculate the power supply drift index.
[0074] The larger the power supply drift index is, the higher the risk of failure of each power data collection node in the power system. The power supply drift index is an important parameter for measuring the stability of the power system. It reflects the degree of deviation of power parameters such as voltage and current from the normal operating state. When the power supply drift index increases, it indicates that the gap between the operating state of the power system and the ideal state is widening, which usually means that there are potential problems or abnormalities in the system.
[0075] First, the increase in the power supply drift index may be caused by equipment aging, load imbalance, or changes in environmental factors. These factors can cause fluctuations and anomalies in parameters such as voltage and current in the power system, increasing the instability of the system. The larger the drift index, the more serious the fluctuations and anomalies, and the higher the probability that the parameters detected by the power data acquisition node deviate from the normal value, thereby increasing the risk of failure.
[0076] Secondly, a large power supply drift index may also indicate that some components in the system are in a critical working state. For example, an excessively large drift index may be caused by overload operation or overheating of the transformer or transmission line. If not handled and adjusted in time, it may cause equipment damage or system crash. When the power data acquisition node works in such an environment, its reliability and accuracy will also be affected, further increasing the possibility of failure.
[0077] Among them, the method for obtaining the communication frequency deviation value is:
[0078] Set a sliding window w of fixed size, send requests to the power system regularly, record the results, and save the information such as the delay and packet loss rate of each command in the sliding window; perform weighted calculation on each command result in the sliding window to obtain the communication stability value hx of the power equipment, and weight each command result according to its time interval in the sliding window. The delay value is Li, the time interval is Ti, and the weight coefficient is Wi, where i represents the result of the i-th command sending. Calculate the weighted communication delay. The specific calculation expression is: Among them, n is the number of command results in the sliding window, yp is the weighted communication delay, the communication stability value and weighted communication delay of the power equipment are analyzed, and the communication frequency deviation value is calculated. The specific calculation expression is: DF = |hx-yp|; where DF is the communication frequency deviation value.
[0079] The larger the communication frequency deviation value, the higher the risk of failure of each power data acquisition node in the power system. The communication frequency deviation value is an important parameter to measure the stability of the communication signal in the power system. It reflects the degree of deviation of the communication frequency between the data acquisition node and the central processing system. When the communication frequency deviation value increases, it indicates that the stability and accuracy of the communication signal are affected, which usually means that there is a problem or interference in the communication link, which in turn affects the reliability of data transmission.
[0080] The increase in communication frequency deviation may be caused by external interference, equipment aging, or improper system configuration. These factors will cause the frequency of the communication link to fluctuate, affecting the stable transmission of the signal. The larger the deviation value, the more serious these fluctuations are, the lower the reliability of the communication signal, and the data transmission between the data acquisition node and the central processing system is frequently interrupted or the bit error increases, thereby increasing the risk of failure.
[0081] A large communication frequency deviation may also indicate performance degradation of the data acquisition node itself or the communication equipment. For example, failure of the sensor or communication module, antenna damage or poor connection can cause significant deviations in the communication frequency. If these problems are not solved in a timely manner, the data acquisition node will not be able to stably and accurately transmit the collected power parameters, resulting in the central processing system being unable to obtain real-time and effective data, increasing the overall system failure risk.
[0082] The power supply drift index and the communication frequency deviation value are normalized, and the risk coefficient of failure of each power data acquisition node in the power system is calculated based on the normalized power supply drift index and the communication frequency deviation value.
[0083] For example, the present invention can use the following formula to calculate the risk coefficient of failure of each power data acquisition node in the power system, and the calculation expression is: Wherein, PN is the risk coefficient, pk is the power supply drift index, DF is the communication frequency deviation value, a1 and a2 are the proportional coefficients of the power supply drift index and the communication frequency deviation value, and a2>a1>0.
[0084] The risk classification module divides the risk of failure of the power data collection nodes in the power system into high-risk collection nodes and low-risk collection nodes according to the evaluation results.
[0085] The acquired risk coefficient of failure of each power data acquisition node in the power system is compared with the risk coefficient reference threshold value. If the risk coefficient of failure of each power data acquisition node in the power system is greater than or equal to the risk coefficient reference threshold value, the risk of failure of the power data acquisition node in the power system is classified as a high-risk acquisition node, and an early warning signal is generated at this time;
[0086] If the risk coefficient of failure of each power data acquisition node in the power system is less than the risk coefficient reference threshold, the risk of failure of the power data acquisition node in the power system is classified as a low-risk acquisition node, and no warning signal is generated at this time.
[0087] The control module performs fault diagnosis on high-risk acquisition nodes, automatically sends out alarm signals and executes emergency treatment measures when a fault is detected.
[0088] In the power system, the system will conduct real-time fault diagnosis for those power data collection nodes that are identified as high-risk. When these nodes are detected to have faults, the system will automatically send out an alarm signal and immediately execute the preset emergency treatment measures to ensure the safe and stable operation of the power system.
[0089] For high-risk nodes, the system will conduct more frequent and detailed monitoring and analysis. For example, the system will collect parameters such as voltage, current, temperature, humidity, etc. in real time and compare them with the preset normal range. The collected data will be analyzed at multiple levels to identify possible fault characteristics, such as voltage imbalance, current harmonics, temperature anomalies, etc. Once data anomalies are detected, the system will immediately determine whether the node is faulty.
[0090] When the risk of failure of a power data acquisition node in the power system is classified as a high-risk acquisition node, that is, the risk coefficient of failure of each power data acquisition node in the power system within a period of time is greater than or equal to the risk coefficient reference threshold, the risk coefficients greater than or equal to the risk coefficient reference threshold in a subsequent period of time are collected and a corresponding data set is established.
[0091] For each high-risk node n∈Nhigh, n is the number of risk coefficients in the set, and Nhigh is the risk coefficient set of high-risk acquisition nodes;
[0092] Collect its power parameter data Dn in real time and perform fault diagnosis. If the fault characteristic function F(Dn) is greater than or equal to the fault threshold TF, it is determined that node n has a fault.
[0093] When a fault is detected at node n, the system generates an alarm signal An and sends the fault information to relevant personnel. An = alarm signal (n)
[0094] According to the fault type Fn and the preset emergency treatment plan E, the corresponding emergency treatment measures are executed, and the emergency treatment function E(Fn) is defined. The system executes corresponding measures according to the fault type Fn.
[0095] In this embodiment, several power data acquisition nodes are arranged in the power system to collect power parameters of the power system in real time; after multi-level analysis of the collected power parameters by the convolutional neural network model; the transmitted power parameters are preprocessed, the fault characteristics in the power parameters are extracted and a fault prediction model is constructed, and the risk of failure of each power data acquisition node in the power system is evaluated; according to the evaluation results, the risk of failure of the power data acquisition nodes in the power system is divided into high-risk acquisition nodes and low-risk acquisition nodes; fault diagnosis is performed on the high-risk acquisition nodes, and an alarm signal is automatically issued and emergency treatment measures are executed when a fault is detected. It can improve the accuracy and timeliness of fault detection, reduce the maintenance cost of the power system, enhance the stability and reliability of the system, effectively prevent and quickly respond to power failures, and ensure the safe operation of the power system.
[0096] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0097] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0098] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0099] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0100] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0101] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0102] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0103] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0104] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.
[0105] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.
Claims
1. Power supply fault monitoring system based on artificial intelligence, Features: It includes data acquisition module, data transmission module, data processing module, risk classification module and control module; A data acquisition module is arranged at several power data acquisition nodes in the power system and is used to collect power parameters of the power system in real time; The data transmission module performs multi-level analysis on the collected power parameters through the convolutional neural network model and transmits them to the data processing module; The data processing module pre-processes the transmitted power parameters, extracts the fault characteristics in the power parameters and builds a fault prediction model, and evaluates the risk of failure of each power data collection node in the power system; A risk classification module, which divides the risk of failure of power data collection nodes in the power system into high-risk collection nodes and low-risk collection nodes according to the evaluation results; The control module performs fault diagnosis on high-risk acquisition nodes, automatically sends out alarm signals and executes emergency treatment measures when a fault is detected.
2. The artificial intelligence-based power supply fault monitoring system according to claim 1, Features: In the data transmission module, the collected power parameters are analyzed at multiple levels through the convolutional neural network model, specifically: The real-time collected power parameters are preprocessed and then input into the trained convolutional neural network model; The convolutional and pooling layers perform multi-level feature extraction on the input data to identify features and patterns in the data; The fully connected layer and the output layer conduct a comprehensive analysis of the extracted features and output the prediction results; Determine whether there is a fault in the power system, as well as the type and location of the fault based on the prediction results; The output results of the convolutional neural network model are transmitted to the data processing module.
3. The power supply fault monitoring system based on artificial intelligence according to claim 2, Features: in, The fault characteristics in the power parameters include the power supply drift index of the power equipment and the communication frequency deviation value of the power equipment.
4. The artificial intelligence-based power supply fault monitoring system according to claim 3, Features: The method for obtaining the power supply drift index is as follows: Obtain the number j of power supply devices in the time period s, and label them as j=1, 2, 3...c, where c is a positive integer greater than 0, calculate the power supply of each power supply device, and calculate the power supply efficiency of each power supply device in the power system; The power supply efficiency of each power supply device within the time period s is established as a set, and the mean of the set, that is, the mean power supply efficiency of the power supply equipment, is calculated. The real-time power supply efficiency of each power supply device is analyzed with the mean power supply efficiency of the power supply equipment, and the standard deviation of the power supply efficiency of all power supply devices in the power system is calculated, that is, the power supply drift index is calculated.
5. The artificial intelligence-based power supply fault monitoring system according to claim 4, Features: The method for obtaining the communication frequency deviation value is: Set a sliding window w of fixed size, send requests to the power system regularly, and record the results, saving information such as the delay and packet loss rate of each command in the sliding window; Perform weighted calculation on each command result in the sliding window to obtain the communication stability value hx of the power equipment. Weight each command result according to its time interval in the sliding window. The delay value is Li, the time interval is Ti, and the weight coefficient is Wi, where i represents the result of the i-th command sending. Calculate the weighted communication delay. The specific calculation expression is: Among them, n is the number of command results in the sliding window, yp is the weighted communication delay, the communication stability value and weighted communication delay of the power equipment are analyzed, and the communication frequency deviation value is calculated. The specific calculation expression is: DF = |hx-yp|; where DF is the communication frequency deviation value.
6. The artificial intelligence-based power supply fault monitoring system according to claim 5, Features: The power supply drift index and the communication frequency deviation value are normalized, and the risk coefficient of failure of each power data acquisition node in the power system is calculated based on the normalized power supply drift index and the communication frequency deviation value.
7. The artificial intelligence-based power supply fault monitoring system according to claim 6, Features: In the risk classification module, the risk of failure of the power data acquisition nodes in the power system is divided into high-risk acquisition nodes and low-risk acquisition nodes, specifically: The acquired risk coefficient of failure of each power data acquisition node in the power system is compared with the risk coefficient reference threshold value. If the risk coefficient of failure of each power data acquisition node in the power system is greater than or equal to the risk coefficient reference threshold value, the risk of failure of the power data acquisition node in the power system is classified as a high-risk acquisition node, and an early warning signal is generated at this time; If the risk coefficient of failure of each power data acquisition node in the power system is less than the risk coefficient reference threshold, the risk of failure of the power data acquisition node in the power system is classified as a low-risk acquisition node, and no warning signal is generated at this time.
8. The artificial intelligence-based power supply fault monitoring system according to claim 7, Features: When the risk of failure of a power data acquisition node in the power system is classified as a high-risk acquisition node, that is, the risk coefficient of failure of each power data acquisition node in the power system within a period of time is greater than or equal to the risk coefficient reference threshold, the risk coefficients greater than or equal to the risk coefficient reference threshold in a subsequent period of time are collected and a corresponding data set is established.
9. The artificial intelligence-based power supply fault monitoring system according to claim 8, Features: For each high-risk node n∈Nhigh, n is the number of risk coefficients in the set, and Nhigh is the risk coefficient set of high-risk acquisition nodes; The power parameter data Dn is collected in real time and fault diagnosis is performed. If the fault characteristic function F(Dn) is greater than or equal to the fault threshold TF, it is determined that node n has a fault. When a fault is detected in node n, the system generates an alarm signal An and sends the fault information to relevant personnel, An = alarm signal (n). According to the fault type Fn and the preset emergency processing plan E, the corresponding emergency processing measures are executed, and the emergency processing function E(Fn) is defined. The system executes corresponding measures according to the fault type Fn.
