Intelligent checking and early warning method for operation risk of distribution network automation equipment and related device

Abstract

The invention relates to the technical field of data processing, and discloses a distribution network automation equipment operation risk intelligent checking and early warning method and a related device, and the method comprises the steps: obtaining a multi-source data set; preprocessing each piece of multi-source data in the multi-source data set to obtain a preprocessed data set; running a risk check model through the target equipment, and performing risk check on each piece of pre-processed data in the pre-processed data set to obtain a risk check data set; operating a risk early-warning model through the target equipment, and performing early-warning analysis on each piece of risk check data in the risk check data set to obtain a risk early-warning data set; and determining and displaying target risk early warning information according to the risk early warning data set. By implementing the method provided by the invention, the accuracy of the operation risk detection of the distribution network automation equipment can be effectively improved, and the purpose of more accurate and more efficient operation risk detection of the distribution network automation equipment can be realized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an intelligent inspection and early warning method for operation risk of distribution network automation equipment and related devices. Background Art

[0002] During the operation of the distribution network, the fault information, hanging / delisting, investment / withdrawal / planning and other status information of the intelligent circuit breaker, the abnormal status information of the pressure plate, the investment / withdrawal / planning and status information of the pressure plate are the basic dispatching information that needs to be paid attention to and mastered in real time during the operation of the distribution network. The accuracy, completeness and effectiveness of this information are related to the stability and safety of the power grid operation arrangement, and directly affect the reliability and other control indicators.

[0003] At present, there is no effective way and means to conduct unified standardized management in this field. It mainly relies on the personal responsibility, personal professional skills and work experience accumulated from long-term work of the distribution network dispatcher to manage and control the operation. The operation results mainly depend on the personal professional quality and professional accomplishment. Due to the large workload and heavy operation tasks of distribution network operation and control, it is difficult to continuously and effectively guarantee the long-term and smooth development and progress of work by relying entirely on individuals. Therefore, it is urgent to develop a set of distribution network automation equipment operation risk inspection and early warning tools with complete functions, simple operation and excellent performance, so as to improve the accuracy of distribution network automation equipment operation risk detection, so as to achieve more accurate and efficient distribution network automation equipment operation risk detection. Summary of the invention

[0004] The present invention provides a distribution network automation equipment operation risk intelligent inspection and early warning method and related devices to solve the technical problem of how to improve the accuracy of distribution network automation equipment operation risk detection.

[0005] In a first aspect, a method for intelligent inspection and early warning of operation risks of distribution network automation equipment is provided, comprising:

[0006] Obtain multi-source data sets;

[0007] Preprocessing each multi-source data in the multi-source data set to obtain a preprocessed data set;

[0008] Running the risk check model on the target device, performing a risk check on each preprocessed data in the preprocessed data set, and obtaining a risk check data set;

[0009] By running the risk warning model on the target device, a warning analysis is performed on each risk inspection data in the risk inspection data set to obtain a risk warning data set;

[0010] Based on the risk warning data set, determine and display the target risk warning information.

[0011] In the second aspect, a distribution network automation equipment operation risk intelligent inspection and early warning device is provided, comprising:

[0012] An acquisition module is used to acquire multi-source data sets;

[0013] A first processing module is used to preprocess each multi-source data in the multi-source data set to obtain a preprocessed data set;

[0014] A second processing module is used to run the risk check model through the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set;

[0015] The third processing module is used to run the risk warning model through the target device, perform warning analysis on each risk inspection data in the risk inspection data set, and obtain the risk warning data set;

[0016] The determination module is used to determine and display target risk warning information based on the risk warning data set.

[0017] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of the above-mentioned distribution network automation equipment operation risk intelligent inspection and early warning method are implemented.

[0018] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned distribution network automation equipment operation risk intelligent inspection and early warning method are implemented.

[0019] In the scheme implemented by the above-mentioned distribution network automation equipment operation risk intelligent inspection and early warning method and related devices, a multi-source data set is obtained, and each multi-source data in the multi-source data set is preprocessed to obtain a preprocessed data set, so as to further perform a risk inspection on each preprocessed data in the preprocessed data set through a target equipment operation risk inspection model to obtain a risk inspection data set, so as to perform early warning analysis on each risk inspection data in the risk inspection data set through a target equipment operation risk early warning model to obtain a risk early warning data set, and then more accurate target risk early warning information can be determined and displayed based on the risk early warning data set to improve the accuracy of distribution network automation equipment operation risk detection, which is conducive to achieving the purpose of more accurate and efficient distribution network automation equipment operation risk detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0021] Figure 1 It is a schematic diagram of an application environment of a distribution network automation equipment operation risk intelligent inspection and early warning method in one embodiment of the present invention;

[0022] Figure 2 It is a flow chart of a method for intelligent inspection and early warning of operation risks of distribution network automation equipment in one embodiment of the present invention;

[0023] Figure 3 It is a process framework diagram of a method for intelligent inspection and early warning of operation risks of distribution network automation equipment in one embodiment of the present invention;

[0024] Figure 4 It is a structural diagram of an intelligent inspection and early warning device for operation risk of distribution network automation equipment in one embodiment of the present invention;

[0025] Figure 5 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0026] Figure 6 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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.

[0028] At present, the most mainstream solution is to manually log into various business systems, such as the Operational Control System (OCS), equipment inventory, outage maintenance plan library, outage management system (OMS) system, etc., to obtain various data through query and statistics, and then manually analyze, count and calculate through electronic spreadsheets, paper forms, etc., and finally form their own customary statistical forms, and then use records, memos, notes, etc. to conduct inspections, early warnings and related operation risk management of distribution network automation equipment.

[0029] Some State Grid Corporation and some power supply bureaus in some provinces and regions of China Southern Grid Corporation have also developed some management and control tools according to their needs, but these tools are developed in the form of product function attached modules. They must either be attached to a certain software or only be associated with a certain manufacturer's hardware or system needs. Once they are separated from the original environment, they will not be able to independently complete the established goals and tasks. Therefore, it is difficult to effectively meet the actual use requirements of other power supply bureaus and cannot be copied and promoted.

[0030] The current mainstream method mainly relies on manual work, and the more prominent shortcomings are as follows: (1) Equipment operation risk inspections are completely dependent on manual work, and the accuracy and completeness of the inspections are difficult to guarantee; (2) Equipment operation risk early warning management is completely dependent on manual work, and the accuracy and stability of early warnings are difficult to effectively guarantee; (3) Distribution network equipment operation risk management mainly relies on manual work, and the effectiveness of management is closely related to the individual's professional and personal qualities. (4) Since relevant data collection, analysis and risk management can only be carried out manually, there is a lack of standardized tool support, and work efficiency is difficult to improve.

[0031] The distribution network automation equipment operation risk intelligent inspection and early warning method provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In an application environment, the client communicates with the server through a network. Exemplarily, taking the operation risk detection scenario for distribution network automation equipment as an example, the target user (R&D personnel or management personnel) can upload a multi-source data set through the client. Accordingly, the server can obtain the multi-source data set through the client, and pre-process each multi-source data in the multi-source data set to obtain a pre-processed data set, so as to further run the risk inspection model through the target device, perform a risk inspection on each pre-processed data in the pre-processed data set, and obtain a risk inspection data set, so that the risk warning model can be run through the target device, and each risk inspection data in the risk inspection data set can be analyzed for warning, and a risk warning data set can be obtained, and then more accurate target risk warning information can be determined and displayed based on the risk warning data set, and the target risk warning information can be fed back to the client. Accordingly, the client can receive the target risk warning information from the server, and can display the target risk warning information on the client for the target user to query or browse. By adopting the intelligent inspection and early warning method for distribution network automation equipment operation risks provided in this application, the accuracy of distribution network automation equipment operation risk detection can be improved, which is conducive to achieving the purpose of more accurate and efficient distribution network automation equipment operation risk detection.

[0032] The client may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices. The server may be implemented by an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0033] See also Figure 2 As shown, Figure 2 A flow chart of a method for intelligent inspection and early warning of operation risk of distribution network automation equipment provided by an embodiment of the present invention includes the following steps:

[0034] S10: Acquire a multi-source data set.

[0035] The multi-source data set may include one or more multi-source data, and the multi-source data may be data from different data sources. For example, the multi-source data may be data from an OCS system, an equipment inventory management system, a power outage maintenance plan library system, etc., and this application does not limit this.

[0036] The server can use multi-source data collection technology to collect multi-source data such as the OCS system, equipment inventory management system, power outage maintenance plan library system, etc., and transmit and aggregate information to obtain the above-mentioned multi-source data set.

[0037] Optionally, a manual solution may be adopted, such as logging into each business system by traditional manual means, and displaying the manually collected, aggregated, analyzed and calculated risk equipment information in physical forms such as electronic spreadsheets, electronic memos, paper notepads, paper notes, paper memos, etc. It is understandable that if the operating status of the equipment changes, the above work and process need to be repeated once, and this application does not limit this.

[0038] S20: Preprocess each multi-source data in the multi-source data set to obtain a preprocessed data set.

[0039] Since the collected multi-source data may have problems such as inconsistent formats and inaccurate data, the server can pre-process the multi-source data, such as cleaning, sorting, and checking, to obtain a pre-processed data set so that subsequent models can perform accurate analysis.

[0040] S30: Running a risk check model on the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set.

[0041] The server can further input the preprocessed data into the target device operation risk inspection model to analyze each preprocessed data through the model to determine whether the device has an operation risk and output a risk inspection data set.

[0042] It should be understood that the process of performing a risk check on each preprocessed data in the preprocessed data set by running the risk check model on the target device to obtain the risk check data set refers to the process of performing a risk check on each preprocessed data in the preprocessed data set by running the risk check model on the target device. In step S20, that is, performing a risk check on each preprocessed data in the preprocessed data set by running the risk check model on the target device to obtain the risk check data set, includes the following steps:

[0043] S21: running a risk inspection model through a target device, performing a read allocation process on each pre-processed data in the pre-processed data set, and obtaining a target allocation data set;

[0044] S22: running a risk check model through the target device, performing a risk check feature engineering operation on each target allocation data in the target allocation data set, and obtaining a risk check feature vector set;

[0045] S23: running a risk check model on the target device, calculating the risk occurrence probability of each risk check feature vector in the risk check feature vector set, and obtaining a risk occurrence probability data set;

[0046] S24: running the risk inspection model through the target device, performing risk result determination and labeling on each risk occurrence probability data in the risk occurrence probability data set, and obtaining a risk inspection data set.

[0047] The server can retrieve each preprocessed data in turn from the preprocessed data set according to the established data storage structure and indexing method to ensure the integrity and accuracy of the data, and perform preliminary verification of the data, such as checking whether the integrity hash value of the data matches and whether the data length meets expectations.

[0048] The server can determine the data distribution method according to the input requirements of the target device operation risk inspection model; if the model requires data to be input in batches, the retrieved pre-processed data can be grouped according to the predetermined batch size. For example, 32 data can be processed in each batch, and 32 consecutive pre-processed data can be grouped together.

[0049] Furthermore, the grouped data can be accurately distributed to the input buffer of the target device running risk inspection model to ensure that the storage format of the data in the buffer is consistent with the format expected by the model, such as converting the data into a specific tensor format or data frame structure, and recording the position index of each data in the buffer to form a target distribution data set. Each element contains the data itself and its index information in the input buffer.

[0050] For each target allocation data in the target allocation data set, its basic features are first extracted. For equipment operation data, it may include but is not limited to the basic parameter features of the equipment, such as the rated power, rated voltage, equipment type and other fixed attribute information of the equipment, as the basic framework for subsequent feature construction. Time-related features of the data can be extracted, such as the timestamp of the data collection time, the day of the week, the time period of the day, and other information; since some equipment may have different operating rules at specific times, this time feature can help discover potential periodic risks.

[0051] Optionally, the equipment operating status characteristics reflected by the target allocation data can be deeply mined. For example, for power equipment, extract statistical characteristics such as the real-time value, average value, maximum value, minimum value, standard deviation, etc. of current and voltage. This status characteristic can reflect the stability and volatility of equipment operation, and there is no restriction on this itself. Calculate the rate of change characteristics of physical quantities such as temperature and pressure at key parts of the equipment, such as the rate of increase or decrease of temperature over a period of time, to determine whether the equipment has abnormal heating or cooling conditions. For equipment with multiple sensors, analyze the correlation characteristics between different sensor data, such as the covariance of the measurement values ​​of two adjacent sensors. Abnormal correlation may indicate problems with the internal structure or connection of the equipment.

[0052] Furthermore, the extracted features can be combined to form an original feature vector. Then, the original feature vector is transformed to improve the expressiveness of the feature and the learning effect of the model. For example, the numerical features are normalized and mapped to a specific interval (such as [0,1] or [-1,1]) to make features of different magnitudes comparable.

[0053] Optionally, for classification features, such as equipment type, etc., one-hot encoding can be used to convert them into numerical vectors so that they can participate in model calculations together with numerical features. Through the above operations, each target allocation data can be converted into a risk check feature vector, and all these vectors constitute a risk check feature vector set.

[0054] Furthermore, each risk check feature vector in the risk check feature vector set may be sequentially input into the target device operation risk check model. The model may process the input feature vector according to its internal algorithm structure and parameters.

[0055] Exemplarily, for a probability model based on machine learning, such as a logistic regression model, the feature vector will be linearly combined with the weight parameters of the model and converted to a risk occurrence probability value between 0 and 1 through a logic function (such as a Sigmoid function). For a deep learning model, such as a neural network, the feature vector will be forward propagated between neurons in each layer, and after feature transformations of multiple hidden layers, the risk occurrence probability value is finally obtained through the activation function of the output layer (such as the normalized exponential function (Softmax), if it is a multi-classification risk situation) or the Sigmoid function (if it is a binary risk situation). All calculated risk occurrence probability values ​​are combined into a risk occurrence probability data set, in which each element corresponds to the risk occurrence probability result of a risk check feature vector.

[0056] The server can set a threshold for the probability of risk occurrence based on the statistical analysis of historical data on equipment operation risks and business requirements. For example, by analyzing the historical failure data and normal operation data of a large number of similar equipment, it is found that when the probability of risk occurrence exceeds 0.6, the possibility of equipment failure in the short term increases significantly, so 0.6 can be set as the risk threshold.

[0057] Furthermore, each risk occurrence probability data in the risk occurrence probability data set is compared with a set threshold value. If the risk occurrence probability is greater than or equal to the threshold value, the device corresponding to the data is determined to be in a high-risk state; if the risk occurrence probability is less than the threshold value, it is determined to be in a low-risk state.

[0058] For situations determined to be high-risk, detailed information on the risk can be further marked, including a preliminary judgment of the risk type. For example, if it is power equipment and abnormal voltage fluctuations lead to an increased risk probability, it can be marked as "abnormal voltage risk"; mark the severity of the risk, such as further dividing it into "high risk-serious" (probability greater than 0.8), "high risk-medium" (probability between 0.6-0.8), etc. according to the size of the probability value; record relevant feature vector information to facilitate subsequent analysis of the cause of the risk, such as marking abnormal voltage values, current change rates and other feature information.

[0059] For low-risk status, some basic information can also be marked, such as "normal operation, no obvious risk signs", and the corresponding feature vector information can be recorded as reference data for the normal operation status of the equipment. The marked risk result information is combined with the corresponding equipment identification, data collection time and other information to form a risk inspection data set, in which each element contains the comprehensive risk inspection result information of the equipment, providing detailed data support for the management and decision-making of equipment operation risks.

[0060] In a possible implementation, the server determines a risk inspection training data set based on the multi-source data set; performs training preprocessing on each risk inspection training data in the risk inspection training data set to obtain a risk inspection training preprocessing data set; constructs an equipment operation risk inspection model based on the multi-source data set based on a convolutional neural network to obtain an initial equipment operation risk inspection model; trains the initial equipment operation risk inspection model based on the risk inspection training preprocessing data set to obtain a reference equipment operation risk inspection model; and tunes the reference equipment operation risk inspection model to obtain a target equipment operation risk inspection model.

[0061] It is understandable that, first, the server can carefully select data closely related to the risk inspection of equipment operation from a rich multi-source data set to determine the risk inspection training data set. The multi-source data may cover various parameters of equipment operation, status information, historical maintenance and environment data, etc. The server can focus on the data subset required for risk inspection through specific screening conditions and business knowledge, and this application does not limit this.

[0062] Next, the server can perform training preprocessing on each data in the risk inspection training data set, which may include but is not limited to data cleaning, such as processing missing values ​​(using mean filling, specific algorithm inference, etc.) and outliers (based on data distribution patterns and business logic judgment and correction), and standardizing the data so that data of different magnitudes and types can participate in model training at the same scale to obtain a risk inspection training preprocessing data set.

[0063] Then, based on the powerful automatic feature extraction and pattern recognition capabilities of convolutional neural networks, the server can build an initial equipment operation risk inspection model according to the characteristics of multi-source data sets and risk inspection task requirements. For example, the server can determine the structure of the network, including the number of convolutional layers, the size and number of convolutional kernels, the settings of the pooling layer, and the architecture of the fully connected layer, and initialize the parameters of the model to give it an initial state to start training.

[0064] Optionally, the server may perform risk inspection feature assessment processing on each multi-source data in the multi-source data set to obtain a risk inspection feature assessment data set; determine the risk inspection convolution layer parameters based on the risk inspection feature assessment data set and the risk inspection requirement data; configure the pooling layer based on the risk inspection convolution layer parameters to obtain the risk inspection pooling layer parameters; construct a fully connected layer based on the risk inspection convolution layer parameters and the risk inspection pooling layer parameters to obtain the risk inspection fully connected layer parameters; based on the convolutional neural network, initialize the model based on the risk inspection convolution layer parameters, the risk inspection pooling layer parameters, the risk inspection fully connected layer parameters and the risk inspection bias parameters to obtain the initial equipment operation risk inspection model.

[0065] That is to say, for each multi-source data in the multi-source data set, the server will carry out risk inspection feature assessment processing. For example, through a series of specific assessment methods and algorithms, the feature information related to risk inspection in each data is deeply mined, and then a risk inspection feature assessment data set is formed. Based on the obtained risk inspection feature assessment data set and the pre-set risk inspection requirement data, the server can further determine the risk inspection convolution layer parameters. For example, determine the key elements such as the size and number of convolution kernels and the step length of the convolution, so as to better capture the risk feature patterns in the data.

[0066] According to the determined risk check convolution layer parameters, the server configures the pooling layer in a targeted manner to obtain the risk check pooling layer parameters. The configuration of the pooling layer is mainly to reduce the data dimension and the amount of calculation, while retaining the main feature information in the data. Based on the previously determined risk check convolution layer parameters and risk check pooling layer parameters, the server can construct a fully connected layer and obtain the risk check fully connected layer parameters. The fully connected layer plays the role of integrating features and making final output judgments in the entire convolutional neural network. Its parameter setting will affect how to effectively fuse and map the features extracted by the convolution layer and the pooling layer to achieve an accurate assessment of the equipment operation risk.

[0067] Based on the convolutional neural network architecture, the server uses the determined risk check convolution layer parameters, risk check pooling layer parameters, risk check full-connection layer parameters and risk check bias parameters to perform a comprehensive initialization operation on the model, thereby obtaining an initial equipment operation risk inspection model. The initial model can have an initial state based on the above-mentioned parameter settings, laying the foundation for subsequent training and optimization, so that it can gradually have the ability to accurately inspect equipment operation risks.

[0068] Furthermore, the server can use the risk inspection training preprocessing data set to train the initial equipment operation risk inspection model. Optionally, the server can divide the data into a training set and a validation set in a certain ratio. During the training process, the model continuously adjusts its weights and parameters based on the training set data to minimize the loss function, and monitors the model performance on the validation set to prevent overfitting and other problems. After multiple rounds of training, the reference equipment operation risk inspection model is obtained.

[0069] The server can tune the reference equipment operation risk inspection model, such as by adjusting hyperparameters (such as learning rate, regularization coefficient, etc.), using regularization techniques (such as L1 / L2 regularization, random dropout, etc.) to optimize the model, or using strategies such as model fusion to combine the advantages of multiple models, thereby obtaining a target equipment operation risk inspection model with better performance, more accuracy and reliability. This model can efficiently and accurately inspect and evaluate equipment operation risks, providing strong support for equipment operation risk management.

[0070] S40: Running the risk warning model on the target device, performing warning analysis on each risk inspection data in the risk inspection data set, and obtaining the risk warning data set.

[0071] Furthermore, the server can input the risk inspection data set into the target device operation risk warning model, so that when the data reaches the warning threshold, a warning can be triggered, thereby obtaining the above-mentioned risk warning data set.

[0072] It should be understood that the process of performing a warning analysis on each risk inspection data set in the risk inspection data set by running the risk warning model of the target device to obtain the risk warning data set refers to the process of performing a warning analysis on each risk inspection data set in the risk inspection data set by running the risk warning model of the target device. Among them, step S40, that is, performing a warning analysis on each risk inspection data set in the risk inspection data set by running the risk warning model of the target device to obtain the risk warning data set, includes the following steps:

[0073] S41: running the risk warning model on the target device, performing data format adaptation processing on each risk inspection data in the risk inspection data set, and obtaining a target format risk inspection data set;

[0074] S42: running a risk warning model on a target device to extract risk warning features from each target format risk inspection data in the target format risk inspection data set to obtain a risk warning feature data set;

[0075] S43: running a risk warning model on the target device to calculate the risk warning probability for each risk warning feature data in the risk warning feature data set, thereby obtaining a risk warning probability data set;

[0076] S44: running the risk warning model through the target device, performing threshold comparison processing on each risk warning probability data in the risk warning probability data set, and obtaining the risk warning data set.

[0077] The server can input the data in the risk inspection data set into the target device operation risk warning model to perform data format adaptation processing for each risk inspection data. Since the data formats from different sources may be different, the model can convert them into the target format that it can process, thereby constructing the target format risk inspection data set to ensure the data standardization and consistency of subsequent operations.

[0078] For each data in the target format risk inspection data set, risk warning features can be further extracted. For example, by using specific feature extraction algorithms and techniques, key features related to equipment operation risk warning in the data can be mined. This feature can reflect the key information of the potential risk of the equipment. After extraction, a risk warning feature data set is formed, which can provide a more targeted data basis for the calculation of risk warning probability.

[0079] Based on the extracted risk warning feature data set, the risk warning probability calculation can be carried out for each risk warning feature data through the model. Optionally, the internal preset warning calculation logic and model algorithm architecture can be used to comprehensively consider various characteristic factors to calculate the probability value of the equipment operation risk corresponding to each data. The probability value can form a risk warning probability data set to quantify the risk level of the equipment.

[0080] Furthermore, a threshold comparison process can be performed for each risk warning probability data in the risk warning probability data set. Optionally, a risk warning threshold can be pre-set. When the calculated risk warning probability exceeds the threshold, it can be determined that the device is in a risk warning state; when the calculated risk warning probability does not exceed the threshold, it can be determined that the device is not in a risk warning state. The relevant data information can be further organized to form a risk warning data set, which may include but is not limited to detailed information on risk warnings, such as risk probability, identification of the equipment involved, etc., so that corresponding risk response measures can be taken in a timely manner based on this information to ensure the safe and stable operation of the equipment.

[0081] It should be noted that the construction method of the target device operation risk warning model can refer to the relevant content of the construction of the aforementioned target device operation risk inspection model, and this application will not go into details here.

[0082] S50: Determine and display target risk warning information according to the risk warning data set.

[0083] The server can determine the risk warning information that needs to be displayed based on the risk warning data set. For example, it can issue warning prompts based on the warning risk type and level, and display them in different colors and texts on the system display screen to facilitate dispatchers to perform management and control operations.

[0084] For example, Figure 3 As shown, Figure 3 The flowchart of the intelligent inspection and early warning method for the operation risk of distribution network automation equipment is shown as an example. It can be understood that this application collects multi-source system data such as OCS system, equipment ledger, power outage inspection plan library system, etc. through multi-source collection, aggregation, and transmission technology, and analyzes and warns the aggregated multi-source data through system construction and built-in equipment operation risk inspection model and equipment operation risk early warning model through built-in inspection and early warning models, and outputs the early warning results to the system display screen. The distribution network dispatcher performs reasonable adjustment and control work based on the risk information and early warning information output by the system, combined with the dispatching procedures and the power outage maintenance plan indicators of Lincang Power Supply Bureau, to ensure the safe and stable operation of the power grid. Optionally, this application is implemented through the use of tools, by manually collecting and analyzing data to form equipment risk data, which is converted into system automation to realize the inspection and early warning of the operation risk of distribution network automation equipment in real time, and these early warning information are displayed on the system, which is convenient for the dispatcher to control the equipment operation risk work in real time.

[0085] Through the implementation of the project, the risk monitoring (i.e. risk inspection) rate of distribution network automation equipment operation can reach 100%, the accuracy of inspection results ≥99.00%, and the inspection speed ≤10 seconds / time; the distribution network risk monitoring (i.e. risk warning) coverage can reach 100%, the accuracy of warning results ≥99.00%, and the real-time warning speed ≤5 minutes / time; the real-time rate of statistical analysis of risk equipment operation can be as high as ≤10 seconds / time; through the use of system tools, the efficiency is improved by 90%, the accuracy is improved by 45%, and the real-time performance is improved by 90%, and this application does not impose any restrictions on this.

[0086] It can be seen that in the above scheme, the server obtains a multi-source data set and pre-processes each multi-source data in the multi-source data set to obtain a pre-processed data set, and further runs a risk check model for the target device to perform a risk check on each pre-processed data in the pre-processed data set to obtain a risk check data set, so that the target device can run a risk warning model to perform a warning analysis on each risk check data in the risk check data set to obtain a risk warning data set, and then more accurate target risk warning information can be determined and displayed based on the risk warning data set to improve the accuracy of risk detection for distribution network automation equipment operation, which is conducive to achieving the purpose of more accurate and efficient risk detection for distribution network automation equipment operation.

[0087] It should be understood that the order of execution of the steps in the above embodiment does not necessarily 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 embodiment of the present invention.

[0088] In one embodiment, a distribution network automation equipment operation risk intelligent inspection and early warning device is provided, and the distribution network automation equipment operation risk intelligent inspection and early warning device corresponds to the distribution network automation equipment operation risk intelligent inspection and early warning method in the above embodiment. Figure 4 As shown, the distribution network automation equipment operation risk intelligent inspection and early warning device includes an acquisition module 101, a first processing module 102, a second processing module 103, a third processing module 104 and a determination module 105. The functional modules are described in detail as follows:

[0089] An acquisition module 101 is used to acquire a multi-source data set;

[0090] A first processing module 102 is used to preprocess each multi-source data in the multi-source data set to obtain a preprocessed data set;

[0091] The second processing module 103 is used to run the risk check model through the target device to perform a risk check on each pre-processed data in the pre-processed data set to obtain a risk check data set;

[0092] The third processing module 104 is used to run the risk warning model through the target device, perform warning analysis on each risk inspection data in the risk inspection data set, and obtain the risk warning data set;

[0093] The determination module 105 is used to determine and display target risk warning information according to the risk warning data set.

[0094] In one embodiment, the second processing module 103 is used to run the risk check model through the target device to perform a risk check on each pre-processed data in the pre-processed data set to obtain a risk check data set, specifically for:

[0095] Running the risk inspection model on the target device, performing read allocation processing on each preprocessed data in the preprocessed data set, and obtaining a target allocation data set;

[0096] Running a risk check model through the target device, performing a risk check feature engineering operation on each target allocation data in the target allocation data set, and obtaining a risk check feature vector set;

[0097] Running a risk check model on the target device, calculating the risk occurrence probability of each risk check feature vector in the risk check feature vector set, and obtaining a risk occurrence probability data set;

[0098] The risk inspection model is run through the target device, and risk result determination and annotation are performed on each risk occurrence probability data in the risk occurrence probability data set to obtain a risk inspection data set.

[0099] In one embodiment, the second processing module 103 is further specifically configured to:

[0100] Determining a risk inspection training data set based on the multi-source data set;

[0101] Performing training preprocessing on each risk inspection training data in the risk inspection training data set to obtain a risk inspection training preprocessing data set;

[0102] Based on a convolutional neural network, an equipment operation risk inspection model is constructed according to the multi-source data set to obtain an initial equipment operation risk inspection model;

[0103] Training the initial equipment operation risk inspection model according to the risk inspection training preprocessing data set to obtain a reference equipment operation risk inspection model;

[0104] The reference equipment operation risk inspection model is tuned to obtain a target equipment operation risk inspection model.

[0105] In one embodiment, the second processing module 103 is used to construct an equipment operation risk inspection model based on the multi-source data set based on a convolutional neural network to obtain an initial equipment operation risk inspection model, including:

[0106] Performing risk inspection feature assessment processing on each multi-source data in the multi-source data set to obtain a risk inspection feature assessment data set;

[0107] Determining risk check convolution layer parameters according to the risk check feature assessment data set and risk check requirement data;

[0108] According to the risk check convolution layer parameters, the pooling layer is configured to obtain risk check pooling layer parameters;

[0109] According to the risk check convolution layer parameters and the risk check pooling layer parameters, a fully connected layer is constructed to obtain risk check fully connected layer parameters;

[0110] Based on the convolutional neural network, the model is initialized according to the risk check convolution layer parameters, the risk check pooling layer parameters, the risk check full-connection layer parameters and the risk check bias parameters to obtain an initial equipment operation risk check model.

[0111] In one embodiment, the third processing module 104 is used to run the risk warning model through the target device to perform warning analysis on each risk inspection data in the risk inspection data set to obtain the risk warning data set, including:

[0112] By running the risk warning model on the target device, each risk inspection data in the risk inspection data set is subjected to data format adaptation processing to obtain a target format risk inspection data set;

[0113] By running the risk warning model on the target device, risk warning features are extracted for each target format risk inspection data in the target format risk inspection data set to obtain a risk warning feature data set;

[0114] By running the risk warning model on the target device, a risk warning probability calculation is performed on each risk warning feature data in the risk warning feature data set to obtain a risk warning probability data set;

[0115] The target device runs a risk warning model, and threshold comparison processing is performed on each risk warning probability data in the risk warning probability data set to obtain a risk warning data set.

[0116] The present invention provides an intelligent inspection and early warning device for operation risks of distribution network automation equipment. The device obtains a multi-source data set and pre-processes each multi-source data in the multi-source data set to obtain a pre-processed data set, and further performs a risk inspection on each pre-processed data in the pre-processed data set through a target equipment operation risk inspection model to obtain a risk inspection data set. Thus, a warning analysis can be performed on each risk inspection data in the risk inspection data set through a target equipment operation risk early warning model to obtain a risk early warning data set. Furthermore, based on the risk early warning data set, more accurate target risk early warning information can be determined and displayed to improve the accuracy of operation risk detection for distribution network automation equipment, which is conducive to achieving the purpose of more accurate and efficient operation risk detection for distribution network automation equipment.

[0117] For the specific definition of the intelligent inspection and early warning device for the operation risk of distribution network automation equipment, please refer to the definition of the intelligent inspection and early warning method for the operation risk of distribution network automation equipment mentioned above, which will not be repeated here. Each module in the above-mentioned intelligent inspection and early warning device for the operation risk of distribution network automation equipment can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0118] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the service side of a distribution network automation equipment operation risk intelligent inspection and early warning method.

[0119] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the client side of a distribution network automation equipment operation risk intelligent inspection and early warning method.

[0120] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0121] Obtain multi-source data sets;

[0122] Preprocessing each multi-source data in the multi-source data set to obtain a preprocessed data set;

[0123] Running the risk check model on the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set;

[0124] By running the risk warning model on the target device, a warning analysis is performed on each risk inspection data in the risk inspection data set to obtain a risk warning data set;

[0125] According to the risk warning data set, target risk warning information is determined and displayed.

[0126] The present invention provides a computer device, which obtains a multi-source data set and pre-processes each multi-source data in the multi-source data set to obtain a pre-processed data set, so as to further perform a risk check on each pre-processed data in the pre-processed data set through a target device operation risk check model to obtain a risk check data set, so as to perform a warning analysis on each risk check data in the risk check data set through a target device operation risk warning model to obtain a risk warning data set, and then more accurate target risk warning information can be determined and displayed based on the risk warning data set to improve the accuracy of distribution network automation equipment operation risk detection, which is conducive to achieving the purpose of more accurate and efficient distribution network automation equipment operation risk detection.

[0127] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0128] Obtain multi-source data sets;

[0129] Preprocessing each multi-source data in the multi-source data set to obtain a preprocessed data set;

[0130] Running the risk check model on the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set;

[0131] By running the risk warning model on the target device, a warning analysis is performed on each risk inspection data in the risk inspection data set to obtain a risk warning data set;

[0132] According to the risk warning data set, target risk warning information is determined and displayed.

[0133] The present invention provides a computer-readable storage medium, which obtains a multi-source data set and pre-processes each multi-source data in the multi-source data set to obtain a pre-processed data set, so as to further perform a risk check on each pre-processed data in the pre-processed data set through a target device operation risk check model to obtain a risk check data set, so as to perform a warning analysis on each risk check data in the risk check data set through a target device operation risk warning model to obtain a risk warning data set, and then more accurate target risk warning information can be determined and displayed based on the risk warning data set to improve the accuracy of distribution network automation equipment operation risk detection, which is conducive to achieving the purpose of more accurate and efficient distribution network automation equipment operation risk detection.

[0134] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0135] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0136] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0137] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A distribution network automation equipment operation risk intelligent inspection and early warning method, characterized in that: The distribution network automation equipment operation risk intelligent inspection and early warning method comprises: Obtain multi-source data sets; Preprocessing each multi-source data in the multi-source data set to obtain a preprocessed data set; Running the risk check model on the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set; By running the risk warning model on the target device, a warning analysis is performed on each risk inspection data in the risk inspection data set to obtain a risk warning data set; According to the risk warning data set, target risk warning information is determined and displayed.

2. The intelligent inspection and early warning method for distribution network automation equipment operation risk according to claim 1 is characterized in that: The risk check model is run on the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set, including: Running the risk inspection model on the target device, performing read allocation processing on each preprocessed data in the preprocessed data set, and obtaining a target allocation data set; Running a risk check model through the target device, performing a risk check feature engineering operation on each target allocation data in the target allocation data set, and obtaining a risk check feature vector set; Running a risk check model on the target device, calculating the risk occurrence probability of each risk check feature vector in the risk check feature vector set, and obtaining a risk occurrence probability data set; The risk inspection model is run through the target device, and risk result determination and annotation are performed on each risk occurrence probability data in the risk occurrence probability data set to obtain a risk inspection data set.

3. The intelligent inspection and early warning method for operation risk of distribution network automation equipment according to claim 1 is characterized in that: The distribution network automation equipment operation risk intelligent inspection and early warning method also includes: Determining a risk inspection training data set based on the multi-source data set; Performing training preprocessing on each risk inspection training data in the risk inspection training data set to obtain a risk inspection training preprocessing data set; Based on a convolutional neural network, an equipment operation risk inspection model is constructed according to the multi-source data set to obtain an initial equipment operation risk inspection model; Training the initial equipment operation risk inspection model according to the risk inspection training preprocessing data set to obtain a reference equipment operation risk inspection model; The reference equipment operation risk inspection model is tuned to obtain a target equipment operation risk inspection model.

4. The intelligent inspection and early warning method for distribution network automation equipment operation risk according to claim 3 is characterized in that: The method of constructing an equipment operation risk inspection model based on the multi-source data set based on the convolutional neural network to obtain an initial equipment operation risk inspection model includes: Performing risk inspection feature assessment processing on each multi-source data in the multi-source data set to obtain a risk inspection feature assessment data set; Determining risk check convolution layer parameters according to the risk check feature assessment data set and risk check requirement data; According to the risk check convolution layer parameters, the pooling layer is configured to obtain risk check pooling layer parameters; According to the risk check convolution layer parameters and the risk check pooling layer parameters, a fully connected layer is constructed to obtain risk check fully connected layer parameters; Based on the convolutional neural network, the model is initialized according to the risk check convolution layer parameters, the risk check pooling layer parameters, the risk check full-connection layer parameters and the risk check bias parameters to obtain an initial equipment operation risk check model.

5. The intelligent inspection and early warning method for operation risk of distribution network automation equipment according to any one of claims 1 to 4, characterized in that: The risk warning model is run on the target device to perform warning analysis on each risk inspection data in the risk inspection data set to obtain the risk warning data set, including: By running the risk warning model on the target device, each risk inspection data in the risk inspection data set is subjected to data format adaptation processing to obtain a target format risk inspection data set; By running the risk warning model on the target device, risk warning features are extracted for each target format risk inspection data in the target format risk inspection data set to obtain a risk warning feature data set; By running the risk warning model on the target device, a risk warning probability calculation is performed on each risk warning feature data in the risk warning feature data set to obtain a risk warning probability data set; The risk warning model is run through the target device, and a threshold comparison process is performed on each risk warning probability data in the risk warning probability data set to obtain a risk warning data set.

6. An intelligent inspection and early warning device for distribution network automation equipment operation risk, characterized in that: The distribution network automation equipment operation risk intelligent inspection and early warning device comprises: An acquisition module is used to acquire multi-source data sets; A first processing module, configured to preprocess each multi-source data in the multi-source data set to obtain a preprocessed data set; A second processing module is used to run a risk check model through a target device to perform a risk check on each pre-processed data in the pre-processed data set to obtain a risk check data set; A third processing module is used to run a risk warning model through a target device to perform a warning analysis on each risk inspection data in the risk inspection data set to obtain a risk warning data set; The determination module is used to determine and display target risk warning information based on the risk warning data set.

7. The intelligent inspection and early warning device for operation risk of distribution network automation equipment according to claim 6 is characterized in that: The second processing module is used to run the risk check model through the target device to perform a risk check on each preprocessed data in the preprocessed data set to obtain a risk check data set, specifically for: Running the risk inspection model on the target device, performing read allocation processing on each preprocessed data in the preprocessed data set, and obtaining a target allocation data set; Running a risk check model through the target device, performing a risk check feature engineering operation on each target allocation data in the target allocation data set, and obtaining a risk check feature vector set; Running a risk check model on the target device, calculating the risk occurrence probability of each risk check feature vector in the risk check feature vector set, and obtaining a risk occurrence probability data set; The risk inspection model is run through the target device, and risk result determination and annotation are performed on each risk occurrence probability data in the risk occurrence probability data set to obtain a risk inspection data set.

8. The intelligent inspection and early warning device for operation risk of distribution network automation equipment according to claim 7 is characterized in that: The second processing module is further specifically configured to: Determining a risk inspection training data set based on the multi-source data set; Performing training preprocessing on each risk inspection training data in the risk inspection training data set to obtain a risk inspection training preprocessing data set; Based on a convolutional neural network, an equipment operation risk inspection model is constructed according to the multi-source data set to obtain an initial equipment operation risk inspection model; Training the initial equipment operation risk inspection model according to the risk inspection training preprocessing data set to obtain a reference equipment operation risk inspection model; The reference equipment operation risk inspection model is tuned to obtain a target equipment operation risk inspection model.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the distribution network automation equipment operation risk intelligent inspection and early warning method as described in any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for intelligent inspection and early warning of operation risks of distribution network automation equipment as described in any one of claims 1 to 5 is implemented.

Images