Power grid power transmission defect intelligent management system based on deep learning convolutional neural network

Abstract

The invention discloses a power grid power transmission defect intelligent management system based on a deep learning convolutional neural network, and belongs to the technical field of power grids, and the system comprises a data collection module which is used for collecting power grid power transmission real-time data; the data processing module is used for processing power grid power transmission real-time data; the model training module is used for constructing a power grid power transmission defect prediction model; the test optimization module is used for performing test optimization on the power grid power transmission defect prediction model; the defect prediction module is used for predicting whether defects exist in the power transmission operation process of the power grid; and the intelligent management module is used for intelligently managing power transmission defects of the power grid. According to the invention, problems that timely prediction and intelligent management cannot be carried out on power transmission defects of a power grid and operation efficiency and safety of the power grid are reduced in the prior art are solved. According to the invention, timely prediction and intelligent management can be carried out on power transmission defects of the power grid based on the deep learning convolutional neural network, potential power transmission fault risks of the power grid can be warned in time, and the operation efficiency and safety performance of the power grid can be improved.

Description

Technical Field

[0001] The present invention relates to the field of power grid technology, and in particular to an intelligent management system for power grid transmission defects based on a deep learning convolutional neural network. Background Art

[0002] With the continuous expansion of my country's power grid and the continuous growth of electricity demand, higher and higher requirements are placed on the safe operation and management of the power grid. During the operation of the power grid, failures such as damage to transmission lines and equipment due to various reasons often occur. These failures will not only affect the normal operation of the power grid, but may also bring safety hazards.

[0003] The Chinese patent with publication number CN115441475A discloses a power emergency control method for power transmission lines. The method extracts power grid status information from the real-time cross-sectional flow file of the power grid, and inputs the power grid status information into the constructed and trained power grid transmission line power control model, outputs the power control action with the highest expected reward value, and obtains the actual reward value generated by executing the control action after the power control action is executed. At the same time, the power control model of the power grid transmission line is updated. The control process does not rely on the power grid model. When the transmission line over-limit problem is detected, the control action and expected effect can be given in time to realize real-time transmission line power control: However, the patent has the following defects:

[0004] The existing deep learning convolutional neural network cannot timely predict and intelligently manage power grid transmission defects, and cannot timely warn of potential power grid transmission failure risks, resulting in poor power grid transmission safety and reliability, and reducing the operating efficiency and safety performance of the power grid. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent management system for power grid transmission defects based on deep learning convolutional neural networks, which can timely predict and intelligently manage power grid transmission defects based on deep learning convolutional neural networks, can timely warn of potential power grid transmission failure risks, can improve the operating efficiency and safety performance of the power grid, and solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The intelligent management system for power transmission defects based on deep learning convolutional neural network includes:

[0008] The data acquisition module is used to collect power grid transmission operation data and power grid transmission environment data, and determine the real-time power grid transmission data based on the Internet of Things;

[0009] A data processing module is used to process the real-time data of power transmission of the power grid based on the Internet of Things and determine the characteristic data of power transmission of the power grid based on the Internet of Things;

[0010] A model training module is used to build a power transmission defect prediction model based on a deep learning convolutional neural network according to the intelligent management requirements of power transmission defects based on a deep learning convolutional neural network.

[0011] A test optimization module is used to test and optimize the power grid transmission defect prediction model based on deep learning convolutional neural network, and determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network;

[0012] A defect prediction module is used to mine and analyze the power transmission characteristic data based on the Internet of Things based on the optimal power transmission defect prediction model based on deep learning convolutional neural network, predict whether there are defects in the power transmission operation process of the power grid, and determine the power transmission defect prediction results based on deep learning convolutional neural network;

[0013] The intelligent management module is used to intelligently manage power grid transmission defects according to the power grid transmission defect intelligent management solution based on deep learning convolutional neural network, so as to ensure the safe operation of power grid transmission.

[0014] Preferably, the data acquisition module comprises:

[0015] The operation collection unit is used to monitor and collect the voltage, current, power and frequency parameters in real time during the power transmission operation of the power grid, and obtain the power transmission operation data of the power grid;

[0016] The environmental collection unit is used to monitor and collect the temperature, humidity and smoke conditions in real time during the power transmission operation of the power grid, and obtain the power transmission environment data of the power grid;

[0017] Among them, based on the power grid transmission operation data and the power grid transmission environment data, the real-time power grid transmission data based on the Internet of Things is determined.

[0018] Preferably, the data processing module includes:

[0019] A data cleaning unit, used to clean the real-time data of power transmission based on the Internet of Things;

[0020] Obtain real-time data on power grid transmission based on the Internet of Things;

[0021] Conduct consistency checks on real-time data of power grid transmission based on IoT;

[0022] According to the reasonable value range and mutual relationship of each parameter in the real-time data of power transmission of power grid based on the Internet of Things, check whether the real-time data of power transmission of power grid based on the Internet of Things meets the requirements;

[0023] Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of power grid transmission based on the Internet of Things;

[0024] Process invalid and missing values ​​of real-time data of power grid transmission based on the Internet of Things;

[0025] According to the intelligent management requirements of power grid transmission defects based on deep learning convolutional neural networks, the real-time data of power grid transmission based on the Internet of Things is checked to determine whether there are invalid values ​​and missing values ​​in the real-time data of power grid transmission based on the Internet of Things;

[0026] Remove invalid and missing values ​​that are of no value to the intelligent management of power grid transmission defects in real-time data of power grid transmission based on the Internet of Things;

[0027] Determine real-time grid transmission data that is valuable for intelligent management of grid transmission defects.

[0028] Preferably, the data processing module further includes:

[0029] A data conversion unit, used for converting the cleaned real-time data of power grid transmission;

[0030] Obtaining real-time data of power grid transmission that is valuable for intelligent management of power grid transmission defects after cleaning;

[0031] Converting valuable real-time data of power grid transmission for intelligent management of power grid transmission defects;

[0032] Eliminate the dimensional differences between real-time data of power grid transmission that are valuable for intelligent management of power grid transmission defects;

[0033] Determine standardized real-time data on power grid transmission;

[0034] A feature extraction unit, used for extracting features from the converted real-time data of power grid transmission;

[0035] Obtaining the converted and standardized real-time data of power grid transmission;

[0036] Extract features from standardized real-time data of power grid transmission;

[0037] Extract features that reflect the intelligent management of power grid transmission defects;

[0038] Determine the power transmission characteristic data of the Internet of Things-based power grid.

[0039] Preferably, the model training module includes:

[0040] A data collection unit, used to collect historical data of power transmission of the Internet of Things-based power grid;

[0041] According to the needs of intelligent management of power grid transmission defects based on deep learning convolutional neural networks, historical data of power grid transmission based on the Internet of Things is collected and reserved;

[0042] A data partitioning unit, used for partitioning the historical data of power transmission of the power grid based on the Internet of Things;

[0043] Obtain historical data on power transmission from the Internet of Things (IoT) grid;

[0044] Divide the historical data of power transmission in the Internet of Things-based power grid;

[0045] Determine a power grid transmission defect prediction training set and a power grid transmission defect prediction test set;

[0046] A model building unit, used to build a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0047] According to the demand for intelligent management of power grid transmission defects based on deep learning convolutional neural network, a deep learning convolutional neural network model framework suitable for intelligent management of power grid transmission defects is selected;

[0048] Based on the power grid transmission defect prediction training set, the selected deep learning convolutional neural network model framework suitable for the intelligent management of power grid transmission defects is trained;

[0049] A power grid transmission defect prediction model based on deep learning convolutional neural network was determined.

[0050] Preferably, the test optimization module includes:

[0051] A test and evaluation unit, used to perform performance test and evaluation on a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0052] Obtain a power grid transmission defect prediction model based on deep learning convolutional neural network;

[0053] Based on the power transmission defect prediction test set, the performance test and evaluation of the power transmission defect prediction model based on deep learning convolutional neural network is carried out;

[0054] Determine the performance test evaluation results based on the power grid transmission defect prediction model;

[0055] An adjustment and optimization unit, used to adjust and optimize parameters of a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0056] According to the performance test evaluation results based on the power grid transmission defect prediction model, the power grid transmission defect prediction model based on deep learning convolutional neural network is mined and analyzed;

[0057] Determine the parameter adjustment optimization scheme based on the power grid transmission defect prediction model;

[0058] According to the parameter adjustment optimization scheme based on the power grid transmission defect prediction model, the parameters of the power grid transmission defect prediction model based on the deep learning convolutional neural network are adjusted and optimized;

[0059] Determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network.

[0060] Preferably, the defect prediction module includes:

[0061] A data extraction unit, used for extracting power transmission characteristic data of the power grid based on the Internet of Things;

[0062] A defect prediction unit is used to predict whether there are defects in the power transmission operation process of the power grid;

[0063] Obtain the optimal power grid transmission defect prediction model based on deep learning convolutional neural network;

[0064] Input the IoT-based power grid transmission characteristic data into the optimal deep learning convolutional neural network-based power grid transmission defect prediction model;

[0065] Based on the optimal power grid transmission defect prediction model based on deep learning convolutional neural network, the power grid transmission characteristic data based on the Internet of Things is mined and analyzed to predict whether there are defects in the power grid transmission operation process, and determine the power grid transmission defect prediction results based on deep learning convolutional neural network.

[0066] Preferably, the intelligent management module includes:

[0067] A plan formulation unit, which is used to formulate an intelligent management plan for power grid transmission defects based on deep learning convolutional neural networks;

[0068] According to the prediction results of power grid transmission defects based on deep learning convolutional neural network, the characteristic data of power grid transmission based on the Internet of Things is mined and analyzed to determine the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network;

[0069] Intelligent management unit, used for intelligent management of power grid transmission defects;

[0070] Obtain intelligent management solutions for power grid transmission defects based on deep learning convolutional neural networks;

[0071] Intelligent management of power grid transmission defects is carried out according to the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network, including:

[0072] The online early warning mechanism is automatically triggered to provide timely early warning of power grid transmission defects, focus on monitoring areas where power grid transmission failures may occur, and intelligently guide operation and maintenance personnel to carry out repairs in the area. At the same time, the real-time status of power grid transmission is shared through wireless communication, allowing operation and maintenance personnel to adjust maintenance plans in a timely manner.

[0073] Preferably, the feature extraction unit comprises:

[0074] Identify modules for:

[0075] The standardized real-time data of power grid transmission is organized into a data matrix, where each row represents the data at one observation time and each column represents a different transmission parameter;

[0076] Compute the covariance matrix of the data matrix;

[0077]

[0078] Where C is the covariance matrix; X is the data matrix; X T is the transposed matrix of X; n is the number of data at the observation time;

[0079] Solve the eigenvalues ​​and eigenvectors of the covariance matrix;

[0080] |C-λI|=0

[0081] (C-λI)v=0

[0082] Where I is the identity matrix; λ is the eigenvalue; v is the eigenvector;

[0083] By analyzing the characteristic vector, the transmission parameters that play a key role in the operation of the power grid are determined as the target transmission parameters, and feature extraction is performed to obtain the first feature set;

[0084] Comparison module for:

[0085] In the first feature set, select a feature and define it as Y, and define other features as G i , G i The data corresponding to and Y are converted into the ranking of their respective column vectors, denoted as R(G i ) and R(Y);

[0086] Calculate the corresponding data R(G in two column vectors i ) and the difference d between R(Y);

[0087]

[0088] Among them, i is each column of feature data; N is the feature length of the data in the first feature set;

[0089] Calculate the correlation coefficient R between two column vectors s ;

[0090]

[0091] The correlation coefficient is compared with a preset threshold, and features with correlation coefficients greater than the preset threshold are extracted from the first feature set to obtain a second feature set, which is used as features that can reflect intelligent management of power grid transmission defects.

[0092] Preferably, the test optimization module further includes:

[0093] The first building block is used to:

[0094] Perform keyword recognition processing on the model parameter text of the power grid transmission defect prediction model before optimization to determine several text keywords;

[0095] Selecting initial configuration keywords from a number of text keywords;

[0096] Evaluate the semantic depth relationship between other text keywords and the initial configuration keywords;

[0097] Taking the initial configuration keyword as the root node, branch expansion is performed according to the semantic depth relationship between each text keyword and the initial configuration keyword. According to the expansion results, several tree subnodes are constructed, and each tree subnode corresponds to an expansion result;

[0098] Constructing a first semantic tree according to a root node and a plurality of tree child nodes;

[0099] A second construction module is used to construct a second semantic tree based on the model parameter text of the optimized power grid transmission defect prediction model;

[0100] Matching modules for:

[0101] Preprocessing the first syntax tree and the second syntax tree to obtain a preprocessed first syntax tree and a preprocessed second syntax tree; the preprocessing includes marking a root node, tree child nodes, and a grammatical relationship between nodes;

[0102] A tree editing operation set is defined for the preprocessed first syntax tree and the preprocessed second syntax tree, respectively, including inserting a node, deleting a node, and converting a node;

[0103] Calculate the tree edit distance between the preprocessed first syntax tree and the preprocessed second syntax tree and the optimal edit operation sequence between each node pair based on a dynamic programming algorithm;

[0104] Establishing a mapping relationship between nodes of the preprocessed first syntax tree and the preprocessed second syntax tree based on the tree edit distance and the optimal edit operation sequence between each node pair;

[0105] According to the node mapping relationship, a plurality of first tree nodes in the first syntax tree are matched with a plurality of second tree nodes in the second syntax tree to determine an incompletely matched node group; a feature analysis is performed on the incompletely matched node group to determine adjustment data of the optimized power grid transmission defect prediction model;

[0106] The adjustment data is compared with the preset adjustment authority data table, and the qualification of the optimized power grid transmission defect prediction model is determined based on the comparison results.

[0107] Compared with the prior art, the present invention has the following beneficial effects:

[0108] The present invention determines the real-time data of power grid transmission based on the Internet of Things by collecting power grid transmission operation data and power grid transmission environment data, determines the power grid transmission characteristic data based on the Internet of Things by processing the real-time data of power grid transmission based on the Internet of Things, constructs a power grid transmission defect prediction model based on deep learning convolutional neural network according to the intelligent management demand of power grid transmission defects based on deep learning convolutional neural network, tests and optimizes the power grid transmission defect prediction model based on deep learning convolutional neural network, determines the optimal power grid transmission defect prediction model based on deep learning convolutional neural network, and determines the optimal power grid transmission defect prediction model based on deep learning convolutional neural network. The network's power grid transmission defect prediction model mines and analyzes the power grid transmission feature data based on the Internet of Things, predicts whether there are defects in the power grid transmission operation process, determines the power grid transmission defect prediction results based on the deep learning convolutional neural network, and intelligently manages the power grid transmission defects according to the power grid transmission defect intelligent management solution based on the deep learning convolutional neural network, so that the power grid transmission can operate safely. It can timely predict and intelligently manage power grid transmission defects based on the deep learning convolutional neural network, and can timely warn of potential power grid transmission failure risks, thereby improving the safety and reliability of power grid transmission and improving the operating efficiency and safety performance of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] Figure 1 This is a structural block diagram of the power grid transmission defect intelligent management system based on deep learning convolutional neural network of the present invention. DETAILED DESCRIPTION

[0110] 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 only 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.

[0111] In order to solve the existing problems of not being able to timely predict and intelligently manage power grid transmission defects based on deep learning convolutional neural networks, not being able to timely warn of potential power grid transmission failure risks, resulting in poor power grid transmission safety and reliability, and reducing the operating efficiency and safety performance of the power grid, please refer to Figure 1 , this embodiment provides the following technical solutions:

[0112] The intelligent management system for power transmission defects in power grid based on deep learning convolutional neural network includes: data acquisition module, data processing module, model training module, test optimization module, defect prediction module and intelligent management module.

[0113] It should be noted that the data acquisition module collects the power grid transmission operation data and the power grid transmission environment data to determine the real-time power grid transmission data based on the Internet of Things; the data processing module processes the real-time power grid transmission data based on the Internet of Things to determine the power grid transmission characteristic data based on the Internet of Things; the model training module constructs a power grid transmission defect prediction model based on a deep learning convolutional neural network; the test optimization module tests and optimizes the power grid transmission defect prediction model based on a deep learning convolutional neural network to determine the optimal power grid transmission defect prediction model based on a deep learning convolutional neural network; the defect prediction module mines and analyzes the power grid transmission characteristic data based on the Internet of Things to predict whether there are defects in the power grid transmission operation process and determine the power grid transmission defect prediction results based on the deep learning convolutional neural network; the intelligent management module intelligently manages the power grid transmission defects to ensure the safe operation of the power grid transmission, and can timely predict and intelligently manage the power grid transmission defects based on the deep learning convolutional neural network, and can timely warn of potential power grid transmission failure risks, thereby improving the safety and reliability of power grid transmission and improving the operation efficiency and safety performance of the power grid.

[0114] Among them, the data acquisition module is used to collect power grid transmission operation data and power grid transmission environment data, and determine the real-time power grid transmission data based on the Internet of Things;

[0115] In this embodiment, the data acquisition module includes:

[0116] The operation collection unit is used to monitor and collect the voltage, current, power and frequency parameters in real time during the power transmission operation of the power grid, and obtain the power transmission operation data of the power grid;

[0117] The environmental collection unit is used to monitor and collect the temperature, humidity and smoke conditions in real time during the power transmission operation of the power grid, and obtain the power transmission environment data of the power grid;

[0118] Among them, based on the power grid transmission operation data and the power grid transmission environment data, the real-time power grid transmission data based on the Internet of Things is determined.

[0119] Among them, the data processing module is used to process the real-time data of power grid transmission based on the Internet of Things and determine the characteristic data of power grid transmission based on the Internet of Things;

[0120] In this embodiment, the data processing module includes:

[0121] A data cleaning unit, used to clean the real-time data of power transmission based on the Internet of Things;

[0122] Obtain real-time data on power grid transmission based on the Internet of Things;

[0123] Conduct consistency checks on real-time data of power grid transmission based on IoT;

[0124] According to the reasonable value range and mutual relationship of each parameter in the real-time data of power transmission of power grid based on the Internet of Things, check whether the real-time data of power transmission of power grid based on the Internet of Things meets the requirements;

[0125] Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of power grid transmission based on the Internet of Things;

[0126] Process invalid and missing values ​​of real-time data of power grid transmission based on the Internet of Things;

[0127] According to the intelligent management requirements of power grid transmission defects based on deep learning convolutional neural networks, the real-time data of power grid transmission based on the Internet of Things is checked to determine whether there are invalid values ​​and missing values ​​in the real-time data of power grid transmission based on the Internet of Things;

[0128] Remove invalid and missing values ​​that are of no value to the intelligent management of power grid transmission defects in real-time data of power grid transmission based on the Internet of Things;

[0129] Identify real-time grid transmission data that is valuable for intelligent management of grid transmission defects;

[0130] A data conversion unit, used for converting the cleaned real-time data of power grid transmission;

[0131] Obtaining real-time data of power grid transmission that is valuable for intelligent management of power grid transmission defects after cleaning;

[0132] Converting valuable real-time data of power grid transmission for intelligent management of power grid transmission defects;

[0133] Eliminate the dimensional differences between real-time data of power grid transmission that are valuable for intelligent management of power grid transmission defects;

[0134] Determine standardized real-time data on power grid transmission;

[0135] A feature extraction unit, used for extracting features from the converted real-time data of power grid transmission;

[0136] Obtaining the converted and standardized real-time data of power grid transmission;

[0137] Extract features from standardized real-time data of power grid transmission;

[0138] Extract features that reflect the intelligent management of power grid transmission defects;

[0139] Determine the power transmission characteristic data of the Internet of Things-based power grid.

[0140] Among them, the model training module is used to build a power grid transmission defect prediction model based on deep learning convolutional neural network according to the intelligent management requirements of power grid transmission defects based on deep learning convolutional neural network;

[0141] In this embodiment, the model training module includes:

[0142] A data collection unit, used to collect historical data of power transmission of the Internet of Things-based power grid;

[0143] According to the needs of intelligent management of power grid transmission defects based on deep learning convolutional neural networks, historical data of power grid transmission based on the Internet of Things is collected and reserved;

[0144] A data partitioning unit, used for partitioning the historical data of power transmission of the power grid based on the Internet of Things;

[0145] Obtain historical data on power transmission from the Internet of Things (IoT) grid;

[0146] Divide the historical data of power transmission in the Internet of Things-based power grid;

[0147] Determine a power grid transmission defect prediction training set and a power grid transmission defect prediction test set;

[0148] A model building unit, used to build a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0149] According to the demand for intelligent management of power grid transmission defects based on deep learning convolutional neural network, a deep learning convolutional neural network model framework suitable for intelligent management of power grid transmission defects is selected;

[0150] Based on the power grid transmission defect prediction training set, the selected deep learning convolutional neural network model framework suitable for the intelligent management of power grid transmission defects is trained;

[0151] A power grid transmission defect prediction model based on deep learning convolutional neural network was determined.

[0152] Among them, the test optimization module is used to test and optimize the power grid transmission defect prediction model based on deep learning convolutional neural network, and determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network;

[0153] In this embodiment, the test optimization module includes:

[0154] A test and evaluation unit, used to perform performance test and evaluation on a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0155] Obtain a power grid transmission defect prediction model based on deep learning convolutional neural network;

[0156] Based on the power transmission defect prediction test set, the performance test and evaluation of the power transmission defect prediction model based on deep learning convolutional neural network is carried out;

[0157] Determine the performance test evaluation results based on the power grid transmission defect prediction model;

[0158] An adjustment and optimization unit, used to adjust and optimize parameters of a power grid transmission defect prediction model based on a deep learning convolutional neural network;

[0159] According to the performance test evaluation results based on the power grid transmission defect prediction model, the power grid transmission defect prediction model based on deep learning convolutional neural network is mined and analyzed;

[0160] Determine the parameter adjustment optimization scheme based on the power grid transmission defect prediction model;

[0161] According to the parameter adjustment optimization scheme based on the power grid transmission defect prediction model, the parameters of the power grid transmission defect prediction model based on the deep learning convolutional neural network are adjusted and optimized;

[0162] Determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network.

[0163] Among them, the defect prediction module is used to mine and analyze the power transmission characteristic data based on the Internet of Things based on the optimal power transmission defect prediction model based on deep learning convolutional neural network, predict whether there are defects in the power transmission operation process of the power grid, and determine the power transmission defect prediction results based on deep learning convolutional neural network;

[0164] In this embodiment, the defect prediction module includes:

[0165] A data extraction unit, used for extracting power transmission characteristic data of the power grid based on the Internet of Things;

[0166] A defect prediction unit is used to predict whether there are defects in the power transmission operation process of the power grid;

[0167] Obtain the optimal power grid transmission defect prediction model based on deep learning convolutional neural network;

[0168] Input the IoT-based power grid transmission characteristic data into the optimal deep learning convolutional neural network-based power grid transmission defect prediction model;

[0169] Based on the optimal power grid transmission defect prediction model based on deep learning convolutional neural network, the power grid transmission characteristic data based on the Internet of Things is mined and analyzed to predict whether there are defects in the power grid transmission operation process, and determine the power grid transmission defect prediction results based on deep learning convolutional neural network.

[0170] Among them, the intelligent management module is used to intelligently manage power grid transmission defects according to the intelligent management solution of power grid transmission defects based on deep learning convolutional neural network, so as to ensure the safe operation of power grid transmission.

[0171] In this embodiment, the intelligent management module includes:

[0172] A plan formulation unit, which is used to formulate an intelligent management plan for power grid transmission defects based on deep learning convolutional neural networks;

[0173] According to the prediction results of power grid transmission defects based on deep learning convolutional neural network, the characteristic data of power grid transmission based on the Internet of Things is mined and analyzed to determine the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network;

[0174] Intelligent management unit, used for intelligent management of power grid transmission defects;

[0175] Obtain intelligent management solutions for power grid transmission defects based on deep learning convolutional neural networks;

[0176] Intelligent management of power grid transmission defects is carried out according to the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network, including:

[0177] The online early warning mechanism is automatically triggered to provide timely early warning of power grid transmission defects, focus on monitoring areas where power grid transmission failures may occur, and intelligently guide operation and maintenance personnel to carry out repairs in the area. At the same time, the real-time status of power grid transmission is shared through wireless communication, allowing operation and maintenance personnel to adjust maintenance plans in a timely manner.

[0178] Preferably, the feature extraction unit comprises:

[0179] Identify modules for:

[0180] The standardized real-time data of power grid transmission is organized into a data matrix, where each row represents the data at one observation time and each column represents a different transmission parameter;

[0181] Compute the covariance matrix of the data matrix;

[0182]

[0183] Where C is the covariance matrix; X is the data matrix; X T is the transposed matrix of X; n is the number of data at the observation time;

[0184] Solve the eigenvalues ​​and eigenvectors of the covariance matrix;

[0185] |C-λI|=0

[0186] (C-λI)v=0

[0187] Where I is the identity matrix; λ is the eigenvalue; v is the eigenvector;

[0188] By analyzing the characteristic vector, the transmission parameters that play a key role in the operation of the power grid are determined as the target transmission parameters, and feature extraction is performed to obtain the first feature set;

[0189] Comparison module for:

[0190] In the first feature set, select a feature and define it as Y, and define other features as G i , G i The data corresponding to and Y are converted into the ranking of their respective column vectors, denoted as R(G i ) and R(Y);

[0191] Calculate the corresponding data R(G in two column vectors i ) and the difference d between R(Y);

[0192]

[0193] Among them, i is each column of feature data; N is the feature length of the data in the first feature set;

[0194] Calculate the correlation coefficient R between two column vectors s ;

[0195]

[0196] The correlation coefficient is compared with a preset threshold, and features with correlation coefficients greater than the preset threshold are extracted from the first feature set to obtain a second feature set, which is used as features that can reflect intelligent management of power grid transmission defects.

[0197] The working principle and beneficial effects of the above technical solution: Transmission parameters are key data describing the electrical characteristics of transmission lines, including line resistance, line inductance, line capacitance, current, voltage, power, etc. In order to improve the efficiency of standardized data processing for standardized real-time power grid transmission data, data is sorted and sorted into a data matrix, in which each row represents data at an observation time and each column represents a different transmission parameter; the covariance matrix of the data matrix is ​​calculated, and the eigenvalues ​​and eigenvectors of the covariance matrix are solved. The eigenvalues ​​are important parameters that reflect the characteristics of the matrix, and they represent the properties of the matrix that remain unchanged during the transformation process. Eigenvectors are vectors corresponding to eigenvalues, which only undergo scaling transformations but not rotations during the matrix transformation process. The positive or negative eigenvalue can reflect certain characteristics of the data, such as positive eigenvalues ​​indicating that the variance of the data in this direction is large, while negative eigenvalues ​​(which usually do not appear in the real number range because the covariance matrix is ​​positive definite) may indicate some abnormal situation or data error. The eigenvector represents the change of the data in a specific direction, which facilitates the understanding of the inherent structure of the data, such as which transmission parameters have a strong correlation. By analyzing the characteristic vector, the transmission parameters that play a key role in the operation of the power grid are determined as the target transmission parameters, thereby providing guidance for the optimization and stable operation of the power system. The target transmission parameters are subjected to feature extraction to obtain the first feature set, which is convenient for avoiding feature extraction for each transmission parameter and improving the efficiency of data feature extraction.

[0198] In this embodiment, in order to perform feature reduction on the first feature set, a feature definition is selected in the first feature set and defined as Y, and other features are defined as G. i , G i The data corresponding to and Y are converted into the ranking of their respective column vectors, the order sequence of the respective data pairs is determined, and the corresponding data R(G i ) and R(Y), and then calculate the correlation R between the two column vectors s , which improves the accuracy of data calculation, and further improves the accuracy of judging the correlation coefficient and the preset threshold value. The preset threshold value is 0.6. The features with correlation coefficients greater than the preset threshold value are extracted from the first feature set to obtain the second feature set, and the second feature set is used as the feature that can reflect the intelligent management of power grid transmission defects. It is convenient to retain the valid data of standardized real-time power grid transmission data to the greatest extent, determine the key features, and improve the efficiency of data processing.

[0199] Preferably, the test optimization module further includes:

[0200] The first building block is used to:

[0201] Perform keyword recognition processing on the model parameter text of the power grid transmission defect prediction model before optimization to determine several text keywords;

[0202] Selecting initial configuration keywords from a number of text keywords;

[0203] Evaluate the semantic depth relationship between other text keywords and the initial configuration keywords;

[0204] Taking the initial configuration keyword as the root node, branch expansion is performed according to the semantic depth relationship between each text keyword and the initial configuration keyword. According to the expansion results, several tree subnodes are constructed, and each tree subnode corresponds to an expansion result;

[0205] Constructing a first semantic tree according to a root node and a plurality of tree child nodes;

[0206] A second construction module is used to construct a second semantic tree based on the model parameter text of the optimized power grid transmission defect prediction model;

[0207] Matching modules for:

[0208] Preprocessing the first syntax tree and the second syntax tree to obtain a preprocessed first syntax tree and a preprocessed second syntax tree; the preprocessing includes marking a root node, tree child nodes, and a grammatical relationship between nodes;

[0209] A tree editing operation set is defined for the preprocessed first syntax tree and the preprocessed second syntax tree, respectively, including inserting a node, deleting a node, and converting a node;

[0210] Calculate the tree edit distance between the preprocessed first syntax tree and the preprocessed second syntax tree and the optimal edit operation sequence between each node pair based on a dynamic programming algorithm;

[0211] Establishing a mapping relationship between nodes of the preprocessed first syntax tree and the preprocessed second syntax tree based on the tree edit distance and the optimal edit operation sequence between each node pair;

[0212] According to the node mapping relationship, a plurality of first tree nodes in the first syntax tree are matched with a plurality of second tree nodes in the second syntax tree to determine an incompletely matched node group; a feature analysis is performed on the incompletely matched node group to determine adjustment data of the optimized power grid transmission defect prediction model;

[0213] The adjustment data is compared with the preset adjustment authority data table, and the qualification of the optimized power grid transmission defect prediction model is determined based on the comparison results.

[0214] The working principle of the above technical solution: In this embodiment, the model parameter text of the power grid transmission defect prediction model before optimization is subjected to keyword recognition processing to determine several text keywords; the model parameter text is text information representing the model parameters. An initial configuration keyword is selected from several text keywords; the initial configuration keyword represents the basic keyword parameters of the model. The semantic depth relationship between other text keywords and the initial configuration keyword is evaluated; the initial configuration keyword is used as the root node (root node), and the tree branches are expanded according to the semantic depth relationship between each text keyword and the initial configuration keyword, and according to the expansion results, several tree subnodes (child nodes) are constructed, each tree subnode corresponds to an expansion result; a first semantic tree is constructed according to the root node and several tree subnodes; it is convenient to organize the model parameter text of the power grid transmission defect prediction model before optimization, form a first semantic tree, and realize the display of the data structure.

[0215] The second construction module constructs a second semantic tree based on the model parameter text of the optimized power grid transmission defect prediction model based on the same method.

[0216] The matching module preprocesses the first syntax tree and the second syntax tree to obtain a preprocessed first syntax tree and a preprocessed second syntax tree; the preprocessing includes marking the root node, the child nodes and the syntax relationship between the nodes; and ensuring that the nodes of the two trees have unique identifiers so as to facilitate comparison and mapping in subsequent steps.

[0217] The tree edit distance between the preprocessed first syntax tree and the preprocessed second syntax tree and the optimal edit operation sequence between each node pair are calculated based on the dynamic programming algorithm, including: Problem definition: Given the preprocessed first syntax tree and the preprocessed second syntax tree, T1 and T2, calculate the minimum number of edit operations required to convert T1 to T2, that is, the tree edit distance. Preorder traversal and array representation: Perform preorder traversal on the two trees T1 and T2 respectively. Store the traversal results in two arrays, recorded as arr1 and arr2 respectively. Dynamic programming array initialization: Define a two-dimensional array dp, where dp[i][j] represents the minimum edit distance required to convert the first i elements of arr1 (representing the partial structure of T1) to the first j elements of arr2 (representing the partial structure of T2). Initialize the dp array: dp[0][0] = 0, indicating that the edit distance from empty tree to empty tree is 0. For i>0, dp[i][0] = i, indicating that it takes i deletion operations to convert the first i elements of T1 (i.e., part of T1) to an empty tree. For j>0, dp[0][j]=j, which means that j insertion operations are required to convert the empty tree to the first j elements of T2 (i.e., part of T2). State transition equation: For the case of i>0 and j>0, perform the following three operations and take the minimum value as the value of dp[i][j]: 1. If the i-th element of arr1 and the j-th element of arr2 are equal: This means that the current node does not need to be edited, so dp[i][j]=dp[i-1][j-1]. If the i-th element of arr1 and the j-th element of arr2 are not equal: One of the following three operations can be performed and the minimum value is taken: Insertion operation: Insert an element equal to the j-th element of arr2 before the i-th element in arr1. This is equivalent to dp[i][j-1]+1 (i.e., on the basis of converting the first i elements of arr1 to the first j-1 elements of arr2, another insertion operation is performed). Delete operation: Delete the i-th element from arr1. This is equivalent to dp[i-1][j]+1 (i.e., on the basis of converting the first i-1 elements of arr1 to the first j elements of arr2, another deletion operation is performed). Replacement operation: Replace the i-th element of arr1 with an element equal to the j-th element of arr2. This is equivalent to dp[i-1][j-1]+1 (i.e., on the basis of converting the first i-1 elements of arr1 to the first j-1 elements of arr2, another replacement operation is performed). Calculate the dp array: Use nested loops to traverse all possible values ​​of i and j, and calculate each element of the dp array according to the above state transition equation. Ultimately, dp[m][n] will represent the minimum edit distance required to convert T1 to T2, where m and n are the lengths of arr1 and arr2, respectively.

[0218] In this embodiment, according to the node mapping relationship, a number of first tree nodes in the first syntax tree are matched with a number of second tree nodes in the second syntax tree to determine an incompletely matched node group; a feature analysis is performed on the incompletely matched node group to determine the adjustment data of the optimized power grid transmission defect prediction model; the adjustment data is compared through a preset adjustment authority data table, and the preset adjustment authority data table presets the model's various parameter adjustment permissions, including maximum value, minimum value, adjustment accuracy, etc. Whether the optimized power grid transmission defect prediction model is qualified is determined based on the comparison result. When it is determined that the adjustment data meets the requirements of the preset adjustment authority data table, it indicates that the optimized power grid transmission defect prediction model is qualified. When it is determined that the adjustment data does not meet the requirements of the preset adjustment authority data table, it indicates that the optimized power grid transmission defect prediction model is unqualified.

[0219] The beneficial effects of the above technical solution are as follows: the model parameter text of the power grid transmission defect prediction model before and after optimization is visualized to obtain the first semantic tree and the second semantic tree. The first semantic tree is structurally compared with the second semantic tree, and the tree editing distance between the preprocessed first syntax tree and the preprocessed second syntax tree and the optimal editing operation sequence between each node pair are calculated based on the dynamic programming algorithm; the mapping relationship between the nodes of the preprocessed first syntax tree and the preprocessed second syntax tree is established; the incompletely matched node group is determined; the feature analysis of the incompletely matched node group is performed to determine the adjustment data of the optimized power grid transmission defect prediction model; the parameter change information of the power grid transmission defect prediction model after optimization relative to that before optimization is accurately determined, and the adjustment data is compared through the preset adjustment authority data table to determine whether the parameter change information meets the requirements, and then accurately determine whether the optimized power grid transmission defect prediction model is qualified.

[0220] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0221] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent management system for power transmission defects based on deep learning convolutional neural network, characterized in that: include: The data acquisition module is used to collect power grid transmission operation data and power grid transmission environment data, and determine the real-time power grid transmission data based on the Internet of Things; A data processing module is used to process the real-time data of power transmission of the power grid based on the Internet of Things and determine the characteristic data of power transmission of the power grid based on the Internet of Things; A model training module is used to build a power transmission defect prediction model based on a deep learning convolutional neural network according to the intelligent management requirements of power transmission defects based on a deep learning convolutional neural network. A test optimization module is used to test and optimize the power grid transmission defect prediction model based on deep learning convolutional neural network, and determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network; A defect prediction module is used to mine and analyze the power transmission characteristic data based on the Internet of Things based on the optimal power transmission defect prediction model based on deep learning convolutional neural network, predict whether there are defects in the power transmission operation process of the power grid, and determine the power transmission defect prediction results based on deep learning convolutional neural network; The intelligent management module is used to intelligently manage power grid transmission defects according to the power grid transmission defect intelligent management solution based on deep learning convolutional neural network, so as to ensure the safe operation of power grid transmission.

2. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 1 is characterized in that: The data acquisition module comprises: The operation collection unit is used to monitor and collect the voltage, current, power and frequency parameters in real time during the power transmission operation of the power grid, and obtain the power transmission operation data of the power grid; The environmental collection unit is used to monitor and collect the temperature, humidity and smoke conditions in real time during the power transmission operation of the power grid, and obtain the power transmission environment data of the power grid; Among them, based on the power grid transmission operation data and the power grid transmission environment data, the real-time power grid transmission data based on the Internet of Things is determined.

3. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 2 is characterized in that: The data processing module comprises: A data cleaning unit, used to clean the real-time data of power transmission based on the Internet of Things; Obtain real-time data on power grid transmission based on the Internet of Things; Conduct consistency checks on real-time data of power grid transmission based on IoT; According to the reasonable value range and mutual relationship of each parameter in the real-time data of power transmission of power grid based on the Internet of Things, check whether the real-time data of power transmission of power grid based on the Internet of Things meets the requirements; Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of power grid transmission based on the Internet of Things; Process invalid and missing values ​​of real-time data of power grid transmission based on the Internet of Things; According to the intelligent management requirements of power grid transmission defects based on deep learning convolutional neural networks, the real-time data of power grid transmission based on the Internet of Things is checked to determine whether there are invalid values ​​and missing values ​​in the real-time data of power grid transmission based on the Internet of Things; Remove invalid and missing values ​​that are of no value to the intelligent management of power grid transmission defects in real-time data of power grid transmission based on the Internet of Things; Determine real-time grid transmission data that is valuable for intelligent management of grid transmission defects.

4. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 3 is characterized in that: The data processing module further includes: A data conversion unit, used for converting the cleaned real-time data of power grid transmission; Obtaining real-time data of power grid transmission that is valuable for intelligent management of power grid transmission defects after cleaning; Converting valuable real-time data of power grid transmission for intelligent management of power grid transmission defects; Eliminate the dimensional differences between real-time data of power grid transmission that are valuable for intelligent management of power grid transmission defects; Determine standardized real-time data on power grid transmission; A feature extraction unit, used for extracting features from the converted real-time data of power grid transmission; Obtaining the converted and standardized real-time data of power grid transmission; Extract features from standardized real-time data of power grid transmission; Extract features that reflect the intelligent management of power grid transmission defects; Determine the power transmission characteristic data of the Internet of Things-based power grid.

5. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 4 is characterized in that: The model training module includes: A data collection unit, used to collect historical data of power transmission of the Internet of Things-based power grid; According to the needs of intelligent management of power grid transmission defects based on deep learning convolutional neural networks, historical data of power grid transmission based on the Internet of Things is collected and reserved; A data partitioning unit, used for partitioning the historical data of power transmission of the power grid based on the Internet of Things; Obtain historical data on power transmission from the Internet of Things (IoT) grid; Divide the historical data of power transmission in the Internet of Things-based power grid; Determine a power grid transmission defect prediction training set and a power grid transmission defect prediction test set; A model building unit, used to build a power grid transmission defect prediction model based on a deep learning convolutional neural network; According to the demand for intelligent management of power grid transmission defects based on deep learning convolutional neural network, a deep learning convolutional neural network model framework suitable for intelligent management of power grid transmission defects is selected; Based on the power grid transmission defect prediction training set, the selected deep learning convolutional neural network model framework suitable for the intelligent management of power grid transmission defects is trained; A power grid transmission defect prediction model based on deep learning convolutional neural network was determined.

6. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 5 is characterized in that: The test optimization module comprises: A test and evaluation unit, used to perform performance test and evaluation on a power grid transmission defect prediction model based on a deep learning convolutional neural network; Obtain a power grid transmission defect prediction model based on deep learning convolutional neural network; Based on the power transmission defect prediction test set, the performance test and evaluation of the power transmission defect prediction model based on deep learning convolutional neural network is carried out; Determine the performance test evaluation results based on the power grid transmission defect prediction model; An adjustment and optimization unit, used to adjust and optimize parameters of a power grid transmission defect prediction model based on a deep learning convolutional neural network; According to the performance test evaluation results based on the power grid transmission defect prediction model, the power grid transmission defect prediction model based on deep learning convolutional neural network is mined and analyzed; Determine the parameter adjustment optimization scheme based on the power grid transmission defect prediction model; According to the parameter adjustment optimization scheme based on the power grid transmission defect prediction model, the parameters of the power grid transmission defect prediction model based on the deep learning convolutional neural network are adjusted and optimized; Determine the optimal power grid transmission defect prediction model based on deep learning convolutional neural network.

7. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 6 is characterized in that: The defect prediction module comprises: A data extraction unit, used for extracting power transmission characteristic data of the power grid based on the Internet of Things; A defect prediction unit is used to predict whether there are defects in the power transmission operation process of the power grid; Obtain the optimal power grid transmission defect prediction model based on deep learning convolutional neural network; Input the IoT-based power grid transmission characteristic data into the optimal deep learning convolutional neural network-based power grid transmission defect prediction model; Based on the optimal power grid transmission defect prediction model based on deep learning convolutional neural network, the power grid transmission characteristic data based on the Internet of Things is mined and analyzed to predict whether there are defects in the power grid transmission operation process, and determine the power grid transmission defect prediction results based on deep learning convolutional neural network.

8. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 7 is characterized in that: The intelligent management module comprises: A plan formulation unit, which is used to formulate an intelligent management plan for power transmission defects based on deep learning convolutional neural networks; According to the prediction results of power grid transmission defects based on deep learning convolutional neural network, the characteristic data of power grid transmission based on the Internet of Things is mined and analyzed to determine the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network; Intelligent management unit, used for intelligent management of power grid transmission defects; Obtain an intelligent management solution for power grid transmission defects based on deep learning convolutional neural networks; Intelligent management of power grid transmission defects is carried out according to the intelligent management scheme of power grid transmission defects based on deep learning convolutional neural network, including: The online early warning mechanism is automatically triggered to provide timely early warning of power grid transmission defects, focus on monitoring areas where power grid transmission failures may occur, and intelligently guide operation and maintenance personnel to carry out repairs in the area. At the same time, the real-time status of power grid transmission is shared through wireless communication, allowing operation and maintenance personnel to adjust maintenance plans in a timely manner.

9. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 4 is characterized in that: The feature extraction unit comprises: Identify modules for: The standardized real-time data of power grid transmission is organized into a data matrix, where each row represents the data at one observation time and each column represents a different transmission parameter; Compute the covariance matrix of the data matrix; Where C is the covariance matrix; X is the data matrix; X T is the transposed matrix of X; n is the number of data at the observation time; Solve the eigenvalues ​​and eigenvectors of the covariance matrix; |C-λI|=0 (C-λI)v=0 Where I is the identity matrix; λ is the eigenvalue; v is the eigenvector; By analyzing the characteristic vector, the transmission parameters that play a key role in the operation of the power grid are determined as the target transmission parameters, and feature extraction is performed to obtain the first feature set; Comparison module for: In the first feature set, select a feature and define it as Y, and define other features as G i , G i The data corresponding to and Y are converted into the ranking of their respective column vectors, denoted as R(G i ) and R(Y); Calculate the corresponding data R(G in two column vectors i ) and the difference d between R(Y); Among them, i is each column of feature data; N is the feature length of the data in the first feature set; Calculate the correlation coefficient R between two column vectors s ; The correlation coefficient is compared with a preset threshold, and features with correlation coefficients greater than the preset threshold are extracted from the first feature set to obtain a second feature set, which is used as features that can reflect intelligent management of power grid transmission defects.

10. The power grid transmission defect intelligent management system based on deep learning convolutional neural network according to claim 6, characterized in that: The test optimization module further includes: The first building block is used to: Perform keyword recognition processing on the model parameter text of the power grid transmission defect prediction model before optimization to determine several text keywords; Selecting initial configuration keywords from a number of text keywords; Evaluate the semantic depth relationship between other text keywords and the initial configuration keywords; Taking the initial configuration keyword as the root node, branch expansion is performed according to the semantic depth relationship between each text keyword and the initial configuration keyword. According to the expansion results, several tree subnodes are constructed, and each tree subnode corresponds to an expansion result; Constructing a first semantic tree according to a root node and a plurality of tree child nodes; A second construction module is used to construct a second semantic tree based on the model parameter text of the optimized power grid transmission defect prediction model; Matching modules for: Preprocessing the first syntax tree and the second syntax tree to obtain a preprocessed first syntax tree and a preprocessed second syntax tree; the preprocessing includes marking a root node, tree child nodes, and a grammatical relationship between nodes; A tree editing operation set is defined for the preprocessed first syntax tree and the preprocessed second syntax tree, respectively, including inserting a node, deleting a node, and converting a node; Calculate the tree edit distance between the preprocessed first syntax tree and the preprocessed second syntax tree and the optimal edit operation sequence between each node pair based on a dynamic programming algorithm; Establishing a mapping relationship between nodes of the preprocessed first syntax tree and the preprocessed second syntax tree based on the tree edit distance and the optimal edit operation sequence between each node pair; According to the node mapping relationship, a plurality of first tree nodes in the first syntax tree are matched with a plurality of second tree nodes in the second syntax tree to determine an incompletely matched node group; a feature analysis is performed on the incompletely matched node group to determine adjustment data of the optimized power grid transmission defect prediction model; The adjustment data is compared with the preset adjustment authority data table, and the qualification of the optimized power grid transmission defect prediction model is determined based on the comparison results.

Images