Power transmission line maintenance method and device, and computer readable program product

Abstract

The invention discloses a maintenance method and device of a power transmission line and a computer readable program product. The method comprises the following steps: receiving operation data of a power transmission line at different sampling time points; performing feature extraction on the operation data to obtain feature parameters of the operation data; inputting the characteristic parameters into a state prediction model, and predicting the line state of the power transmission line based on the characteristic parameters by using the state prediction model to obtain a prediction result; generating a maintenance strategy for the power transmission line according to the prediction result; and maintaining the power transmission line according to the maintenance strategy. According to the invention, the technical problem that the prediction precision is difficult to meet the engineering requirements due to the limitation in multiple aspects when the life cycle of the power transmission line is predicted by using a traditional method in the related technology and the prediction is not carried out by fusing multiple factors is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system operation and maintenance, and in particular, to a method and device for maintaining a power transmission line, and a computer-readable program product. Background Art

[0002] As an important infrastructure for power transmission, the safety and stability of transmission lines are directly related to the stable operation and power supply quality of the power system. However, when using traditional methods to comprehensively predict and manage the life cycle of transmission lines, there are great limitations in data acquisition, analysis and prediction accuracy. Existing methods for predicting transmission lines usually only analyze and predict from a single perspective, and are relatively extensive in data preprocessing and feature extraction. They fail to fully explore and utilize the deep information in the data, and the prediction accuracy and stability cannot meet the high standards required by actual projects.

[0003] The traditional methods used in the above-mentioned related technologies to predict the life cycle of transmission lines have limitations in many aspects and fail to integrate multiple factors for prediction, resulting in the problem that the prediction accuracy is difficult to meet engineering requirements. No effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a transmission line maintenance method and device, and a computer-readable program product, so as to at least solve the technical problem that the traditional method used in the related art to predict the life cycle of the transmission line has limitations in many aspects, fails to integrate multiple factors for prediction, and results in the prediction accuracy being difficult to meet engineering requirements.

[0005] According to one aspect of an embodiment of the present invention, a maintenance method for a transmission line is provided, comprising: receiving operating data of the transmission line at different sampling time points, wherein the operating data is data measured by a plurality of sensors deployed on the transmission line during monitoring of the transmission line, and the plurality of sensors are respectively deployed at different positions on the transmission line; performing feature extraction on the operating data to obtain feature parameters of the operating data; inputting the feature parameters into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the feature parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using a plurality of sets of training data, each of the plurality of sets of training data comprising: sample feature parameters, and sample prediction results corresponding to the sample feature parameters; generating a maintenance strategy for the transmission line according to the prediction result; and maintaining the transmission line according to the maintenance strategy.

[0006] Optionally, receiving operating data of the transmission line at different sampling time points includes at least one of the following: receiving first operating data of the transmission line, wherein the first operating data is data measured during a process of monitoring the transmission line by a vibration sensor deployed on the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of a tower on the transmission line; receiving second operating data of the transmission line, wherein the second operating data is data measured during a process of monitoring the transmission line by a temperature sensor deployed on the transmission line, and the temperature sensor is used to measure a temperature change value of the transmission line under different environmental conditions; receiving third operating data of the transmission line, wherein the third operating data is data measured during a process of monitoring the transmission line by a current sensor deployed on the transmission line, and the current sensor is used to measure a current load change value of the transmission line.

[0007] Optionally, feature extraction is performed on the operation data to obtain feature parameters of the operation data, including: determining the operation data sampled by the same sensor at different sampling time points as an operation data group; performing decentralized processing on each of the operation data groups to obtain a target operation data group, wherein the decentralized processing refers to updating each item of the operation data in the operation data group to the difference between the operation data and the corresponding data mean, and the data mean is the mean of all the operation data in each of the operation data groups; sequentially calculating the covariance between each of the target operation data groups and the initial feature parameters in all the target operation data groups at the same sampling time point to obtain a covariance matrix, wherein the initial feature parameters refer to the difference between the operation data and the corresponding data mean. The difference; analyzing the correlation between two initial feature parameters in the initial feature parameter group according to the covariance matrix to obtain an analysis result, wherein the initial feature parameter group is a combination of the initial feature parameters in two different target operation data groups at the same sampling time point; determining the initial feature parameter group whose correlation is not greater than the correlation threshold as the first initial feature parameter group according to the analysis result, and determining the initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; determining two of the initial feature parameters in the first initial feature parameter group and any one of the initial feature parameter groups in the second initial feature parameter group as target feature parameters; performing dimensionality reduction on the parameter set composed of the target feature parameters to obtain the feature parameters.

[0008] Optionally, sequentially calculating the covariance between each target operation data group and the initial characteristic parameters in all the target operation data groups at the same sampling time point to obtain a covariance matrix includes: sequentially calculating the covariance between each target operation data group and all the target operation data groups using a first formula, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of the sampling time points, X i and Y i respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; after obtaining the covariance between each of the target operation data groups and all the target operation data groups, the covariance is arranged in a matrix form to obtain the covariance matrix.

[0009] Optionally, analyzing the correlation between two initial feature parameters in the initial feature parameter group according to the covariance matrix to obtain an analysis result includes: calculating the correlation between the two initial feature parameters in the initial feature parameter group according to the covariance matrix using a second formula, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance between the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; comparing the correlation with the correlation threshold to obtain a comparison result; and determining the comparison result as the analysis result.

[0010] Optionally, the parameter set composed of the target feature parameters is reduced in dimension to obtain the feature parameters, including: determining a set composed of all the target feature parameters as a first feature parameter set; performing eigendecomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; sorting the eigenvalues ​​in descending order to obtain a sorting result; selecting a predetermined number of the eigenvectors as principal component eigenvectors in order according to the sorting result; constructing a transformation matrix based on the principal component eigenvectors, wherein the transformation matrix is ​​used to perform feature dimensionality reduction on the first feature parameter set; using the transformation matrix to transform the first feature parameter set to obtain a second feature parameter set after the first feature parameter set is subjected to feature dimensionality reduction; and determining the target feature parameters in the second feature parameter set as the feature parameters.

[0011] Optionally, before inputting the characteristic parameters into the state prediction model to use the state prediction model to predict the current state of the transmission line based on the characteristic parameters and obtaining the prediction result, the transmission line maintenance method also includes: obtaining multiple groups of sample data sets, wherein each of the multiple groups of sample data sets is a data set after feature dimensionality reduction; dividing the multiple groups of sample data sets to obtain sample training data sets and sample test data sets, wherein the sample training data sets are used to train the state prediction model, and the sample test data sets are used to test the state prediction model; using the sample training data sets to iteratively train the support vector machine model until the first iteration termination condition is reached, determining that the support vector machine model obtained by the current training is the initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches the maximum number of iterations, and the prediction error does not exceed the error threshold; selecting the sample training data sets except the target sample training data set. The initial state prediction model is trained using the sample training data set outside the training set, and the initial state prediction model is iteratively cross-validated using the target sample training data set until a second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple groups of sample training data sets, and the second iteration termination condition is: each of the multiple groups of sample training data sets has been used as the target sample training data set to verify the initial state prediction model; when the first verification result indicates that the prediction error of the initial state prediction model does not exceed the error threshold, the accuracy of the initial state prediction model is verified using the sample test data set to obtain a second verification result; when the second verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold, the initial state prediction model is determined to be the state prediction model.

[0012] Optionally, the transmission line maintenance method also includes: in the process of iteratively training the support vector machine model using the sample training data set, setting an initial candidate value of a regularization parameter, wherein the regularization parameter represents the degree of penalty for violation intervals, and the violation interval refers to the distance between a sample point in the sample training data set and a decision boundary being less than a predetermined interval; in the process of iteratively cross-validating the initial state prediction model using the sample training data set, updating the initial candidate value using a grid search method until an update termination condition is reached, determining that the current candidate value is the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches a preset model state, and the model state refers to the state of the initial state prediction model in terms of the prediction error and the accuracy.

[0013] According to another aspect of an embodiment of the present invention, there is also provided a maintenance device for a transmission line, comprising: a receiving unit, for receiving operating data of the transmission line at different sampling time points, wherein the operating data is data measured in the process of monitoring the transmission line by multiple sensors deployed on the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line; a first acquisition unit, for extracting features from the operating data to obtain characteristic parameters of the operating data; a second acquisition unit, for inputting the characteristic parameters into a state prediction model, so as to predict the line state of the transmission line based on the characteristic parameters using the state prediction model to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each of the multiple sets of training data includes: sample characteristic parameters, and sample prediction results corresponding to the sample characteristic parameters; a generation unit, for generating a maintenance strategy for the transmission line according to the prediction result; and a maintenance unit, for maintaining the transmission line according to the maintenance strategy.

[0014] Optionally, the receiving unit comprises at least one of the following: a first receiving module for receiving first operating data of the transmission line, wherein the first operating data is data measured during the process of monitoring the transmission line by a vibration sensor deployed on the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of the tower on the transmission line; a second receiving module for receiving second operating data of the transmission line, wherein the second operating data is data measured during the process of monitoring the transmission line by a temperature sensor deployed on the transmission line, and the temperature sensor is used to measure the temperature change value of the transmission line under different environmental conditions; a third receiving module for receiving third operating data of the transmission line, wherein the third operating data is data measured during the process of monitoring the transmission line by a current sensor deployed on the transmission line, and the current sensor is used to measure the current load change value of the transmission line.

[0015] Optionally, the first acquisition unit includes: a first determination module, used to determine that the operation data sampled by the same sensor at different sampling time points are an operation data group; a first acquisition module, used to perform decentralized processing on each of the operation data groups to obtain a target operation data group, wherein the decentralized processing refers to updating each of the operation data in the operation data group to the difference between the operation data and the corresponding data mean, and the data mean is the mean of all the operation data in each of the operation data groups; a second acquisition module, used to sequentially calculate the covariance between each of the target operation data groups and the initial feature parameters in all the target operation data groups at the same sampling time point to obtain a covariance matrix, wherein the initial feature parameters refer to the difference between the operation data and the corresponding data mean; a third acquisition module, used to The correlation between two initial feature parameters in the initial feature parameter group is analyzed according to the covariance matrix to obtain an analysis result, wherein the initial feature parameter group is a combination of the initial feature parameters in two different target operation data groups at the same sampling time point; a second determination module is used to determine, according to the analysis result, the initial feature parameter group whose correlation is not greater than the correlation threshold as the first initial feature parameter group, and to determine the initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; a third determination module is used to determine that two of the initial feature parameters in the first initial feature parameter group and any one of the initial feature parameter groups in the second initial feature parameter group are target feature parameters; a fourth acquisition module is used to perform dimensionality reduction on the parameter set composed of the target feature parameters to obtain the feature parameters.

[0016] Optionally, the second acquisition module includes: a first calculation submodule, configured to sequentially calculate the covariance between each target operation data group and all target operation data groups using a first formula, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of the sampling time points, X i and Y i respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; a first acquisition submodule is used to arrange the covariance in a matrix form after acquiring the covariance between each target operation data group and all the target operation data groups to obtain the covariance matrix.

[0017] Optionally, the third acquisition module includes: a second calculation submodule, used to calculate the correlation between two initial feature parameters in the initial feature parameter group using a second formula according to the covariance matrix, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance between the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; a second acquisition submodule, used to compare the correlation with the correlation threshold to obtain a comparison result; a first determination submodule, used to determine that the comparison result is the analysis result.

[0018] Optionally, the fourth acquisition module includes: a second determination submodule, used to determine that a set consisting of all the target feature parameters is a first feature parameter set; a third acquisition submodule, used to perform eigendecomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; a fourth acquisition submodule, used to sort the eigenvalues ​​in descending order to obtain a sorting result; a selection submodule, used to select a predetermined number of the eigenvectors as principal component eigenvectors in sequence according to the sorting result; a construction submodule, used to construct a transformation matrix based on the principal component eigenvectors, wherein the transformation matrix is ​​used to perform feature dimension reduction on the first feature parameter set; a fifth acquisition submodule, used to transform the first feature parameter set using the transformation matrix to obtain a second feature parameter set after the first feature parameter set is subjected to feature dimension reduction; the third determination submodule, used to determine that the target feature parameters in the second feature parameter set are the feature parameters.

[0019] Optionally, the transmission line maintenance device also includes: a third acquisition unit, used to obtain multiple groups of sample data sets before inputting the characteristic parameters into the state prediction model to predict the current state of the transmission line based on the characteristic parameters using the state prediction model to obtain the prediction results, wherein each of the multiple groups of sample data sets is a data set after feature dimensionality reduction; a fourth acquisition unit, used to divide the multiple groups of sample data sets to obtain sample training data sets and sample test data sets, wherein the sample training data sets are used to train the state prediction model, and the sample test data sets are used to test the state prediction model; a first determination unit, used to iteratively train the support vector machine model using the sample training data sets until a first iteration termination condition is reached, and determine that the support vector machine model obtained by the current training is the initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches the maximum number of iterations, and the prediction error does not exceed the error threshold; a fifth acquisition unit, used to select the sample training data set except the target state prediction model. The initial state prediction model is trained by the sample training data set other than the target sample training data set, and the initial state prediction model is iteratively cross-validated using the target sample training data set until a second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple groups of sample training data sets, and the second iteration termination condition is that each of the multiple groups of sample training data sets has been used as the target sample training data set to verify the initial state prediction model; a sixth acquisition unit is used to verify the accuracy of the initial state prediction model using the sample test data set to obtain a second verification result when the first verification result indicates that the prediction error of the initial state prediction model does not exceed the error threshold; a second determination unit is used to determine that the initial state prediction model is the state prediction model when the second verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold.

[0020] Optionally, the maintenance device of the transmission line also includes: a third determination unit, used to set an initial candidate value of a regularization parameter during the process of iteratively training the support vector machine model using the sample training data set, wherein the regularization parameter represents the degree of penalty for violation intervals, and the violation interval refers to the distance between the sample point in the sample training data set and the decision boundary being less than a predetermined interval; a fourth determination unit, used to update the initial candidate value using a grid search method during the process of iterative cross-validation of the initial state prediction model using the sample training data set, until an update termination condition is reached, and determine that the current candidate value is the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches a preset model state, and the model state refers to the state of the initial state prediction model in terms of the prediction error and the accuracy.

[0021] According to another aspect of an embodiment of the present invention, a transmission line maintenance system is further provided. The transmission line maintenance system uses any one of the transmission line maintenance methods described above.

[0022] According to another aspect of an embodiment of the present invention, there is further provided a computer-readable storage medium, the computer-readable storage medium comprising a stored program, wherein the program executes any one of the above-mentioned transmission line maintenance methods.

[0023] According to another aspect of an embodiment of the present invention, a processor is further provided, wherein the processor is used to run a program, wherein the program executes any one of the above-mentioned transmission line maintenance methods when running.

[0024] According to another aspect of an embodiment of the present invention, a computer program product is provided, comprising computer instructions, wherein when the computer instructions are executed by a processor, any one of the above-mentioned transmission line maintenance methods is executed.

[0025] In an embodiment of the present invention, operation data of a transmission line at different sampling time points are received, wherein the operation data is data measured by multiple sensors deployed on the transmission line during monitoring of the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line; feature extraction is performed on the operation data to obtain feature parameters of the operation data; the feature parameters are input into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the feature parameters to obtain prediction results, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each set of the multiple sets of training data includes: sample feature parameters, and sample prediction results corresponding to the sample feature parameters; a maintenance strategy for the transmission line is generated according to the prediction results; and the transmission line is maintained according to the maintenance strategy. Through the above technical scheme, the purpose of collecting multi-source data by using multiple sensors and analyzing and processing the data collected by the sensors by using the state prediction model is achieved, so as to predict the current state and remaining life of the transmission line, and take corresponding measures to maintain the transmission line according to the prediction results. The technical effect of integrating multiple factors and using the model to predict the state of the transmission line is achieved, and the prediction accuracy is improved, thereby solving the technical problem that the traditional methods used in related technologies to predict the life cycle of the transmission line have limitations in many aspects, and the prediction fails to integrate multiple factors, resulting in the prediction accuracy being difficult to meet the engineering requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0027] Figure 1 It is a hardware structure block diagram of a mobile terminal of a method for maintaining a power transmission line according to an embodiment of the present invention;

[0028] Figure 2 is a flow chart of a method for maintaining a power transmission line according to an embodiment of the present invention;

[0029] Figure 3 is a flow chart of an optional transmission line maintenance method according to an embodiment of the present invention;

[0030] Figure 4 Schematic diagram of a transmission line maintenance device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] The following is an explanation of terms that may be involved in the embodiments of the present invention.

[0034] PCA (Principal Component Analysis) is a statistical method used to discover patterns in a data set. It converts the data into a new coordinate system through orthogonal transformation to maximize the variance of any projection of the data. This method is often used in areas such as dimensionality reduction, data compression, and pattern recognition.

[0035] SVM (Support Vector Machine) is a supervised learning algorithm mainly used for classification and regression tasks. It distinguishes different categories by finding the optimal boundary between data points (a straight line in two-dimensional space and a hyperplane in higher-dimensional space).

[0036] As introduced in the background technology, the traditional methods used in the related art to predict the life cycle of power transmission lines have limitations in many aspects, and do not integrate multiple factors for prediction, resulting in prediction accuracy that is difficult to meet engineering requirements. In view of the above defects, a maintenance method and device for power transmission lines, and a computer-readable program product are provided in the embodiments of the present invention.

[0037] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0038] The method embodiments provided in the embodiments of the present invention can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for maintaining a power transmission line according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0039] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the maintenance method of the power transmission line in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above network include but are not limited to the Internet, corporate intranets, local area networks, mobile communication networks and combinations thereof. The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0040] According to an embodiment of the present invention, a method embodiment of a transmission line maintenance method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0041] Figure 2 is a flow chart of a method for maintaining a power transmission line according to an embodiment of the present invention, such as Figure 2 As shown, the method comprises the following steps:

[0042] Step S202, receiving operation data of the transmission line at different sampling time points, wherein the operation data is data measured by multiple sensors deployed on the transmission line during monitoring of the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line.

[0043] Combine the following Figure 3 The above embodiments of the present invention are described in detail. Figure 3 is a flow chart of an optional transmission line maintenance method according to an embodiment of the present invention.

[0044] As shown in Figure 3, a variety of sensors can be deployed at key locations on the transmission line to collect the operating data of the transmission line in real time. Each sensor can then transmit the operating data collected in real time to a central data platform via wireless communication technology. The central data platform can analyze and process the operating data collected by each sensor after receiving it, so as to analyze and monitor the health status of the transmission line.

[0045] The key parts here usually refer to specific points or areas that have a significant impact on the health and safety of the transmission line. For transmission lines, these key parts may include: 1) Pole towers or electric poles: Pole towers are the infrastructure that supports the transmission line, and their stability directly affects the safety of the line; for example, deploying vibration sensors on pole towers can monitor structural vibrations and promptly detect problems such as structural fatigue and looseness; temperature sensors can detect temperature changes of pole towers under different environmental conditions and evaluate the impact of thermal expansion and contraction; current sensors are used to monitor current changes passing through pole towers to determine whether there is overload or abnormal current; 2) Wire and cable connection points: The wire and cable connection points of the transmission line are susceptible to stress, oxidation, overheating, etc., and are high-incidence areas of failure; for example, temperature sensors can To monitor the temperature of the connection point to determine whether there is an overheating problem; the current sensor is used to monitor the current passing through the connection point to evaluate its electrical performance; 3) Insulators: Insulators are important components to prevent current leakage. Deterioration in their performance will increase the risk of line short circuits; for example, temperature sensors can detect the temperature of insulators during operation to determine whether they have overheating or aging problems; current sensors can monitor the leakage current of insulators to determine their insulation performance; 4) Key electrical equipment such as transformers and circuit breakers: The operating status of these devices is crucial to the health of transmission lines; for example, vibration sensors can monitor the vibration of equipment to determine whether it is running smoothly; temperature sensors are used to monitor equipment temperature to evaluate its load and cooling conditions; current sensors monitor current fluctuations to determine the electrical condition of the equipment.

[0046] Step S204: extract features from the operating data to obtain feature parameters of the operating data.

[0047] As above Figure 3 As shown, the central data platform can clean, preprocess and extract features of the operating data transmitted by each sensor to obtain characteristic parameters of the line operation status.

[0048] Step S206, inputting the characteristic parameters into the state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the characteristic parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each of the multiple sets of training data includes: sample characteristic parameters, and sample prediction results corresponding to the sample characteristic parameters.

[0049] As above Figure 3As shown, a transmission line life prediction model (i.e., the above-mentioned state prediction model) can be established based on characteristic parameters by applying a machine learning model, and the established transmission line life prediction model can be used to monitor the health status of the transmission line in real time and generate prediction results; when establishing the state prediction model here, characteristic parameters extracted from operating data in a historical time period can be used for training and testing, and in the process of using the trained state prediction model to predict the health status of the transmission line based on its operating data, the model can also perform adaptive learning to continuously optimize and improve model performance.

[0050] Step S208: generating a maintenance strategy for the transmission line according to the prediction result.

[0051] As above Figure 3 As shown, a maintenance plan and maintenance strategy for the transmission line can be generated according to the prediction results of the above state prediction model.

[0052] Step S210: Maintain the transmission line according to the maintenance strategy.

[0053] In this embodiment, the transmission line may be maintained according to the maintenance strategy obtained in the above step S208 to improve the line quality of the transmission line, extend the service life of the transmission line, etc.

[0054] As can be seen from the above, through the technical solution provided by the above embodiment of the present invention, the operation data of the transmission line at different sampling time points can be received, wherein the operation data is data measured by multiple sensors deployed on the transmission line in the process of monitoring the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line; feature extraction is performed on the operation data to obtain feature parameters of the operation data; the feature parameters are input into the state prediction model to predict the line state of the transmission line based on the feature parameters using the state prediction model to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each of the multiple sets of training data includes: sample feature parameters, sample prediction results corresponding to the sample feature parameters; a maintenance strategy for the transmission line is generated according to the prediction result; and the transmission line is maintained according to the maintenance strategy, so as to achieve the purpose of collecting multi-source data by using multiple sensors, and analyzing and processing the data collected by the sensors by using the state prediction model to predict the current state and remaining life of the transmission line, and taking corresponding measures to maintain the transmission line according to the prediction results, so as to achieve the technical effect of integrating multiple factors and using the model to predict the state of the transmission line, and improve the prediction accuracy.

[0055] Therefore, the technical solution provided by the above-mentioned embodiment of the present invention solves the technical problem that the traditional method used in the related art to predict the life cycle of the transmission line has limitations in many aspects and fails to integrate multiple factors for prediction, resulting in the prediction accuracy being difficult to meet engineering requirements.

[0056] According to the above embodiment of the present invention, receiving operation data of a transmission line at different sampling time points includes at least one of the following: receiving first operation data of the transmission line, wherein the first operation data is data measured during a process of monitoring the transmission line by a vibration sensor deployed on the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of a tower on the transmission line; receiving second operation data of the transmission line, wherein the second operation data is data measured during a process of monitoring the transmission line by a temperature sensor deployed on the transmission line, and the temperature sensor is used to measure a temperature change value of the transmission line under different environmental conditions; receiving third operation data of the transmission line, wherein the third operation data is data measured during a process of monitoring the transmission line by a current sensor deployed on the transmission line, and the current sensor is used to measure a current load change value of the transmission line.

[0057] Specifically, vibration sensors, temperature sensors and current sensors can be installed at key locations of the transmission line, respectively, so that these sensors can be used to monitor and collect the operating data of the transmission line in real time, and the operating data collected in real time can be transmitted to the central data platform through wireless communication technology. After receiving the operating data collected by each sensor, the central data platform can analyze and process the data to analyze and monitor the health status of the transmission line.

[0058] According to the above embodiment of the present invention, feature extraction is performed on the operation data to obtain feature parameters of the operation data, including: determining the operation data sampled by the same sensor at different sampling time points as an operation data group; performing decentralized processing on each operation data group to obtain a target operation data group, wherein the decentralized processing refers to updating each operation data in the operation data group to the difference between the operation data and the corresponding data mean, and the data mean is the mean of all operation data in each operation data group; sequentially calculating the covariance between each target operation data group and the initial feature parameters in all target operation data groups at the same sampling time point to obtain a covariance matrix, wherein the initial feature parameters refer to the difference between the operation data and the corresponding data mean. The difference between them; analyzing the correlation of two initial feature parameters in the initial feature parameter group according to the covariance matrix to obtain an analysis result, wherein the initial feature parameter group is a combination of initial feature parameters in two different target operation data groups at the same sampling time point; determining the initial feature parameter group whose correlation is not greater than the correlation threshold as the first initial feature parameter group according to the analysis result, and determining the initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; determining the two initial feature parameters in the first initial feature parameter group and any one initial feature parameter group in the second initial feature parameter group as target feature parameters; performing dimensionality reduction on the parameter set composed of the target feature parameters to obtain feature parameters.

[0059] Specifically, the preprocessed data can be organized into a matrix, in which each row represents the measurement value of each sensor at a certain time point, and each column represents the measurement reading of a certain sensor at each time point. For each column of sensor readings in the matrix, its mean is calculated to form a mean vector (i.e., the above-mentioned data mean), and the corresponding mean is subtracted from the measurement value of each time point of the sensor readings in each column to obtain a centralized data set. The covariance is calculated based on the data set obtained after decentralized processing, and a covariance matrix is ​​formed. The correlation between the features (initial feature parameters) is identified by analyzing the covariance matrix, and feature selection and feature dimensionality reduction are performed to screen out the feature parameters in the operating data; feature selection is performed by correlation analysis to deduplicate the features with high correlation in the initial feature parameters, and feature dimensionality reduction can reduce the dimension of the data while retaining the key information in the data set, so as to improve the efficiency of data analysis.

[0060] In the above embodiment of the present invention, the covariance between each target operation data group and the initial characteristic parameters in all target operation data groups at the same sampling time point is calculated in sequence to obtain a covariance matrix, including: using a first formula to calculate the covariance between each target operation data group and all target operation data groups in sequence, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of sampling time points, X i and Y i Respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; after obtaining the covariance between each target operation data group and all target operation data groups, the covariance is arranged in a matrix form to obtain a covariance matrix.

[0061] Specifically, we can use the formula Calculate the covariance to form a covariance matrix, where X and Y represent the initial feature parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of sampling time points, and X i and Y i Respectively represent the operating data of different sensors at the i-th sampling time point, and They respectively represent the data means corresponding to two different sensors; after obtaining the covariance between each target operation data group and all target operation data groups, the covariance is arranged in a matrix form to obtain a covariance matrix; here the covariance matrix represents the correlation coefficient between the target operation data group corresponding to each sensor and the target operation data groups corresponding to all sensors including itself, the data on the diagonal represents the correlation between each sensor and itself, and the data in the target operation data group here is the data set obtained after decentralized processing of the readings of each sensor.

[0062] In the above embodiment of the present invention, the correlation between two initial feature parameters in the initial feature parameter group is analyzed according to the covariance matrix to obtain an analysis result, including: calculating the correlation between the two initial feature parameters in the initial feature parameter group according to the covariance matrix using the second formula, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance of the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; comparing the correlation with the correlation threshold to obtain a comparison result; and determining the comparison result as the analysis result.

[0063] Specifically, the formula r(X,Y)=Cov(X,Y) / σXσY can be used to calculate the correlation between two features in the feature pair (two initial feature parameters in the initial feature parameter group), where r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance of the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter. The calculated correlation is compared with the correlation threshold to analyze the correlation between the features. If the correlation exceeds the correlation threshold, it is considered that the features are positively correlated, that is, the two features are highly correlated and similar, and one of them can be retained; if the correlation does not exceed the correlation threshold, it is considered that the features are negatively correlated, that is, the two features are less correlated, and no deduplication is required, thereby avoiding the problem of multicollinearity, reducing the similarity and complexity of the data, and improving the efficiency of data analysis.

[0064] In the above embodiment of the present invention, the parameter set composed of target feature parameters is reduced in dimension to obtain feature parameters, including: determining a set composed of all target feature parameters as a first feature parameter set; performing feature decomposition on a covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; sorting the eigenvalues ​​in descending order to obtain a sorting result; selecting a predetermined number of eigenvectors as principal component eigenvectors in order according to the sorting result; constructing a conversion matrix based on the principal component eigenvectors, wherein the conversion matrix is ​​used to perform feature dimension reduction on the first feature parameter set; using the conversion matrix to transform the first feature parameter set to obtain a second feature parameter set after the feature dimension reduction of the first feature parameter set; and determining the target feature parameters in the second feature parameter set as feature parameters.

[0065] Specifically, after deduplication of feature pairs with high similarity (i.e., deduplication of any one of them), feature dimensionality reduction processing can be performed through principal component analysis (PCA): first, the remaining initial feature parameters can be used as target feature parameters to form a first feature parameter set, and the covariance matrix corresponding to the first feature parameter set is subjected to eigendecomposition to obtain eigenvalues ​​and eigenvectors; then, according to the size of the eigenvalue, the eigenvectors corresponding to the first k largest eigenvalues ​​can be selected as principal components, and then a transformation matrix is ​​constructed using the selected principal component eigenvectors, and finally, the features of the first feature parameter set are transformed through the PCA transformation matrix to obtain a reduced-dimensional data set (i.e., a second feature parameter set), and the target feature parameters in the second feature parameter set are the feature parameters selected after feature dimensionality reduction.

[0066] According to the above embodiment of the present invention, before inputting the characteristic parameters into the state prediction model to predict the current state of the transmission line based on the characteristic parameters using the state prediction model and obtaining the prediction result, the transmission line maintenance method also includes: obtaining multiple groups of sample data sets, wherein each of the multiple groups of sample data sets is a data set after feature dimensionality reduction; dividing the multiple groups of sample data sets to obtain sample training data sets and sample test data sets, wherein the sample training data sets are used to train the state prediction model, and the sample test data sets are used to test the state prediction model; using the sample training data sets to iteratively train the support vector machine model until the first iteration termination condition is reached, determining that the support vector machine model obtained by the current training is the initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches the maximum number of iterations, and the prediction error does not exceed the error threshold; selecting the sample training data The sample training data sets other than the target sample training data set are concentrated to train the initial state prediction model, and the target sample training data set is used to iteratively cross-validate the initial state prediction model until the second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple sample training data sets, and the second iteration termination condition is: each of the multiple sample training data sets has been used as the target sample training data set to verify the initial state prediction model; when the first verification result indicates that the prediction error of the initial state prediction model does not exceed the error threshold, the sample test data set is used to verify the accuracy of the initial state prediction model to obtain a second verification result; when the second verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold, the initial state prediction model is determined to be the state prediction model.

[0067] Specifically, the operation data of the transmission line in the historical time period can be obtained, and after feature selection and feature dimension reduction processing, the data is used as a sample data set to iteratively train the support vector machine model (SVM) to preliminarily obtain a state prediction model (i.e., the above-mentioned initial state prediction model). Then, the K-fold cross-validation method can be used to iteratively cross-validate the performance of the initial state prediction model. When the result of the iterative cross-validation indicates that the prediction error of the initial state prediction model meets the accuracy requirement (i.e., the prediction error does not exceed the error threshold), the root mean square error can be used to perform performance verification on the current initial state prediction model in the sample test data set. If the verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold, the initial state prediction model can be considered as a state prediction model for use. The K-fold cross-validation method here is a method for evaluating model performance and optimizing model parameters. The data set is divided into K subsets, and K-1 subsets are used to train the model each time, and the remaining subset is used to verify the model performance. This process is repeated K times, and each verification set is different. Finally, the average performance of the K verifications is calculated as the evaluation result of the model. Through this process, the performance of the model on unseen data can be more accurately evaluated.

[0068] According to the above embodiment of the present invention, the maintenance method of the transmission line also includes: in the process of iteratively training the support vector machine model using the sample training data set, setting the initial candidate value of the regularization parameter, wherein the regularization parameter represents the degree of penalty for the violation interval, and the violation interval refers to the distance between the sample point in the sample training data set and the decision boundary is less than the predetermined interval; in the process of iteratively cross-validating the initial state prediction model using the sample training data set, updating the initial candidate value using the grid search method until the update termination condition is reached, determining the current candidate value as the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches the preset model state, and the model state refers to the state of the initial state prediction model in terms of prediction error and accuracy.

[0069] Specifically, in the process of iteratively training the support vector machine model using a sample training data set, the initial candidate value of the regularization parameter C can be set. The regularization parameter C has a significant impact on the complexity of the model and the degree of fit to the training data. By setting its initial value, you can begin to explore how this parameter affects model performance. The regularization parameter C is an important hyperparameter in SVM, which controls the degree of penalty for violating the interval. A smaller C value means that the model has a higher tolerance for interval violations and may produce a hyperplane with a larger interval, that is, the model complexity is lower and may tend to underfit; a larger C value means that the model has a lower tolerance for interval violations and may produce a hyperplane with a smaller interval, that is, the model complexity is higher and may tend to overfit. Setting the initial candidate value of C is the first step in model parameter adjustment. In the process of iterative cross-validation of the performance of the initial state prediction model using the K-fold cross-validation method, in each fold of cross-validation, the model will make predictions on the validation set and calculate the prediction error, such as accuracy, recall, F1 score, root mean square error (RMSE), etc., to evaluate the performance of the model. Through multiple cross-validations, the average performance indicators of the model under a set of C parameter values ​​can be obtained. Then, by changing the value of the C parameter and repeating the K-fold cross validation, we can find the C value that can optimize the model performance index (for example, minimize RMSE) to find the optimal C value. After finding the optimal C value, we use the entire training set data and this optimal C value to retrain the model to obtain the final SVM model with the best generalization ability (that is, the state prediction model obtained in the above steps).

[0070] It should be noted that when setting the initial candidate value of the regularization parameter C, a more appropriate initial candidate value can be selected based on experience or relevant text records, or the initial candidate value can be set by random initialization or by giving a heuristic rule. There is no specific restriction on what value is set as the initial candidate value of the regularization parameter C.

[0071] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0072] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0073] According to an embodiment of the present invention, there is also provided a transmission line maintenance device for implementing the above-mentioned transmission line maintenance method. Figure 4 Schematic diagram of a maintenance device for a power transmission line according to an embodiment of the present invention. Figure 4 As shown, the device comprises: a receiving unit 41, a first obtaining unit 43, a second obtaining unit 45, a generating unit 47 and a maintaining unit 49. The following is a detailed description of the transmission line maintenance device.

[0074] The receiving unit 41 is used to receive the operation data of the transmission line at different sampling time points, wherein the operation data is data measured by multiple sensors deployed on the transmission line during the process of monitoring the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line.

[0075] The first acquisition unit 43 is used to extract features from the operation data to obtain feature parameters of the operation data.

[0076] The second acquisition unit 45 is used to input the characteristic parameters into the state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the characteristic parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each set of the multiple sets of training data includes: sample characteristic parameters, and sample prediction results corresponding to the sample characteristic parameters.

[0077] The generating unit 47 is used to generate a maintenance strategy for the transmission line according to the prediction result.

[0078] The maintenance unit 49 is used to maintain the transmission line according to the maintenance strategy.

[0079] It should be noted here that the above-mentioned receiving unit 41, first acquisition unit 43, second acquisition unit 45, generation unit 47 and maintenance unit 49 correspond to steps S202 to S210 in the above-mentioned embodiments, and the five units are the same as the instances and application scenarios implemented by the corresponding steps, but are not limited to the contents disclosed in the above-mentioned embodiments.

[0080] As can be seen from the above, in the scheme recorded in the above-mentioned embodiment of the present invention, a receiving unit can be used to receive the operation data of the transmission line at different sampling time points, wherein the operation data is data measured in the process of monitoring the transmission line by multiple sensors deployed on the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line; then a first acquisition unit is used to extract features of the operation data to obtain feature parameters of the operation data; then a second acquisition unit is used to input the feature parameters into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the feature parameters to obtain a prediction result, wherein the state prediction model is a machine learning model using multiple sets of training data. The plurality of training data are obtained by training, and each of the plurality of training data includes: sample feature parameters, and sample prediction results corresponding to the sample feature parameters; then the generation unit is used to generate a maintenance strategy for the transmission line according to the prediction results; finally, the maintenance unit is used to maintain the transmission line according to the maintenance strategy, thereby achieving the purpose of collecting multi-source data by using multiple sensors, analyzing and processing the data collected by the sensors by using a state prediction model, so as to predict the current state and remaining life of the transmission line, and taking corresponding measures to maintain the transmission line according to the prediction results, realizing the technical effect of integrating multiple factors and using the model to predict the state of the transmission line, and improving the prediction accuracy.

[0081] Therefore, the technical solution provided by the above-mentioned embodiment of the present invention solves the technical problem that the traditional method used in the related art to predict the life cycle of the transmission line has limitations in many aspects and fails to integrate multiple factors for prediction, resulting in the prediction accuracy being difficult to meet engineering requirements.

[0082] Optionally, the receiving unit comprises at least one of the following: a first receiving module for receiving first operating data of the transmission line, wherein the first operating data is data measured during the process of monitoring the transmission line by a vibration sensor deployed on the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of the tower on the transmission line; a second receiving module for receiving second operating data of the transmission line, wherein the second operating data is data measured during the process of monitoring the transmission line by a temperature sensor deployed on the transmission line, and the temperature sensor is used to measure the temperature change value of the transmission line under different environmental conditions; a third receiving module for receiving third operating data of the transmission line, wherein the third operating data is data measured during the process of monitoring the transmission line by a current sensor deployed on the transmission line, and the current sensor is used to measure the current load change value of the transmission line.

[0083] Optionally, the first acquisition unit includes: a first determination module, which is used to determine that the operation data sampled by the same sensor at different sampling time points are the operation data group; a first acquisition module, which is used to perform decentralized processing on each operation data group to obtain a target operation data group, wherein the decentralized processing refers to updating each operation data in the operation data group to the difference between the operation data and the corresponding data mean, and the data mean is the mean of all operation data in each operation data group; a second acquisition module, which is used to sequentially calculate the covariance between each target operation data group and the initial feature parameters in all target operation data groups at the same sampling time point to obtain a covariance matrix, wherein the initial feature parameters refer to the difference between the operation data and the corresponding data mean; a third acquisition module, which is used to obtain the covariance matrix according to The covariance matrix analyzes the correlation between two initial feature parameters in the initial feature parameter group to obtain an analysis result, wherein the initial feature parameter group is a combination of initial feature parameters in two different target operation data groups at the same sampling time point; the second determination module is used to determine, according to the analysis result, an initial feature parameter group whose correlation is not greater than a correlation threshold as the first initial feature parameter group, and to determine an initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; the third determination module is used to determine two initial feature parameters in the first initial feature parameter group and any one initial feature parameter group in the second initial feature parameter group as target feature parameters; the fourth acquisition module is used to reduce the dimension of the parameter set composed of the target feature parameters to obtain feature parameters.

[0084] Optionally, the second acquisition module includes: a first calculation submodule, used to sequentially calculate the covariance between each target operation data group and all target operation data groups using a first formula, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of sampling time points, Xi and Y i Respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; the first acquisition submodule is used to arrange the covariance in a matrix form after acquiring the covariance between each target operation data group and all target operation data groups to obtain a covariance matrix.

[0085] Optionally, the third acquisition module includes: a second calculation submodule, used to calculate the correlation between two initial feature parameters in the initial feature parameter group using a second formula according to the covariance matrix, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance of the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; a second acquisition submodule, used to compare the correlation with a correlation threshold to obtain a comparison result; and a first determination submodule, used to determine that the comparison result is an analysis result.

[0086] Optionally, the fourth acquisition module includes: a second determination submodule, used to determine that the set consisting of all target feature parameters is a first feature parameter set; a third acquisition submodule, used to perform eigendecomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; a fourth acquisition submodule, used to sort the eigenvalues ​​in descending order to obtain a sorting result; a selection submodule, used to select a predetermined number of eigenvectors as principal component eigenvectors in sequence according to the sorting result; a construction submodule, used to construct a transformation matrix based on the principal component eigenvectors, wherein the transformation matrix is ​​used to perform feature dimension reduction on the first feature parameter set; a fifth acquisition submodule, used to transform the first feature parameter set using the transformation matrix to obtain a second feature parameter set after the first feature parameter set is subjected to feature dimension reduction; the third determination submodule, used to determine that the target feature parameters in the second feature parameter set are feature parameters.

[0087] Optionally, the transmission line maintenance device also includes: a third acquisition unit, which is used to obtain multiple groups of sample data sets before inputting the characteristic parameters into the state prediction model to predict the current state of the transmission line based on the characteristic parameters using the state prediction model, and obtaining the prediction results, wherein each of the multiple groups of sample data sets is a data set after feature dimensionality reduction; a fourth acquisition unit, which is used to divide the multiple groups of sample data sets to obtain sample training data sets and sample test data sets, wherein the sample training data sets are used to train the state prediction model, and the sample test data sets are used to test the state prediction model; a first determination unit, which is used to iteratively train the support vector machine model using the sample training data sets until the first iteration termination condition is reached, and determine that the support vector machine model obtained by the current training is the initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches the maximum number of iterations, and the prediction error does not exceed the error threshold; a fifth acquisition unit, which is used to select The initial state prediction model is trained with sample training data sets other than the target sample training data set in the sample training data set, and the initial state prediction model is iteratively cross-validated using the target sample training data set until a second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple groups of sample training data sets, and the second iteration termination condition is that each of the multiple groups of sample training data sets has been used as the target sample training data set to verify the initial state prediction model; a sixth acquisition unit is used to verify the accuracy of the initial state prediction model using the sample test data set when the first verification result indicates that the prediction error of the initial state prediction model does not exceed the error threshold, and obtain a second verification result; a second determination unit is used to determine that the initial state prediction model is the state prediction model when the second verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold.

[0088] Optionally, the maintenance device of the transmission line also includes: a third determination unit, used to set the initial candidate value of the regularization parameter during the iterative training of the support vector machine model using the sample training data set, wherein the regularization parameter represents the degree of penalty for violation interval, and the violation interval refers to the distance between the sample point in the sample training data set and the decision boundary being less than a predetermined interval; a fourth determination unit, used to update the initial candidate value using a grid search method during the iterative cross-validation of the initial state prediction model using the sample training data set, until the update termination condition is reached, and determine that the current candidate value is the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches a preset model state, and the model state refers to the state of the initial state prediction model in terms of prediction error and accuracy.

[0089] According to another aspect of an embodiment of the present invention, a transmission line maintenance system is further provided. The transmission line maintenance system uses any one of the above-mentioned transmission line maintenance methods.

[0090] According to another aspect of an embodiment of the present invention, there is further provided a computer-readable storage medium, the computer-readable storage medium comprising a stored program, wherein the program executes any one of the above-mentioned transmission line maintenance methods.

[0091] Optionally, in this embodiment, the computer-readable storage medium may be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the communication devices in a communication device group.

[0092] Optionally, in this embodiment, a computer-readable storage medium is configured to store program codes for executing the following steps: receiving operating data of a transmission line at different sampling time points, wherein the operating data is data measured in a process of monitoring the transmission line by multiple sensors deployed on the transmission line, and the multiple sensors are respectively deployed at different positions on the transmission line; performing feature extraction on the operating data to obtain feature parameters of the operating data; inputting the feature parameters into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the feature parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each set of the multiple sets of training data includes: sample feature parameters, and sample prediction results corresponding to the sample feature parameters; generating a maintenance strategy for the transmission line according to the prediction result; and maintaining the transmission line according to the maintenance strategy.

[0093] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: receiving first operating data of the transmission line, wherein the first operating data is data measured during the process of monitoring the transmission line by a vibration sensor deployed on the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of the tower on the transmission line; receiving second operating data of the transmission line, wherein the second operating data is data measured during the process of monitoring the transmission line by a temperature sensor deployed on the transmission line, and the temperature sensor is used to measure the temperature change value of the transmission line under different environmental conditions; receiving third operating data of the transmission line, wherein the third operating data is data measured during the process of monitoring the transmission line by a current sensor deployed on the transmission line, and the current sensor is used to measure the current load change value of the transmission line.

[0094] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: determining that the operating data sampled by the same sensor at different sampling time points are operating data groups; performing decentralized processing on each operating data group to obtain a target operating data group, wherein the decentralized processing refers to updating each operating data item in the operating data group to the difference between the operating data and the corresponding data mean, and the data mean is the mean of all operating data in each operating data group; sequentially calculating the covariance between each target operating data group and the initial feature parameters in all target operating data groups at the same sampling time point to obtain a covariance matrix, wherein the initial feature parameters refer to the difference between the operating data and the corresponding data mean. The difference between the values; analyzing the correlation between two initial feature parameters in the initial feature parameter group according to the covariance matrix to obtain an analysis result, wherein the initial feature parameter group is a combination of initial feature parameters in two different target operation data groups at the same sampling time point; determining the initial feature parameter group whose correlation is not greater than the correlation threshold as the first initial feature parameter group according to the analysis result, and determining the initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; determining the two initial feature parameters in the first initial feature parameter group and any one initial feature parameter group in the second initial feature parameter group as target feature parameters; performing dimensionality reduction on the parameter set composed of the target feature parameters to obtain feature parameters.

[0095] Optionally, in this embodiment, the computer-readable storage medium is configured to store a program code for executing the following steps: sequentially calculating the covariance between each target operation data group and all target operation data groups using a first formula, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of sampling time points, X i and Y i Respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; after obtaining the covariance between each target operation data group and all target operation data groups, the covariance is arranged in a matrix form to obtain a covariance matrix.

[0096] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: calculating the correlation between two initial feature parameters in the initial feature parameter group using a second formula according to the covariance matrix, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance of the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; comparing the correlation with the correlation threshold to obtain a comparison result; and determining the comparison result as the analysis result.

[0097] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for executing the following steps: determining a set consisting of all target feature parameters as a first feature parameter set; performing eigendecomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; sorting the eigenvalues ​​in descending order to obtain a sorting result; selecting a predetermined number of eigenvectors in order as principal component eigenvectors according to the sorting result; constructing a transformation matrix based on the principal component eigenvectors, wherein the transformation matrix is ​​used to perform feature dimension reduction on the first feature parameter set; transforming the first feature parameter set using the transformation matrix to obtain a second feature parameter set after the first feature parameter set is subjected to feature dimension reduction; and determining the target feature parameters in the second feature parameter set as feature parameters.

[0098] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: obtaining multiple groups of sample data sets, wherein each of the multiple groups of sample data sets is a data set after feature dimensionality reduction; dividing the multiple groups of sample data sets to obtain sample training data sets and sample test data sets, wherein the sample training data sets are used to train the state prediction model, and the sample test data sets are used to test the state prediction model; iteratively training the support vector machine model using the sample training data sets until a first iteration termination condition is reached, determining that the support vector machine model obtained by the current training is the initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches the maximum number of iterations, and the prediction error does not exceed the error threshold; selecting a sample training data set other than the target sample training data set in the sample training data set The initial state prediction model is trained using a target sample training data set, and the initial state prediction model is iteratively cross-validated using a target sample training data set until a second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple sample training data sets, and the second iteration termination condition is: each of the multiple sample training data sets has been used as a target sample training data set to verify the initial state prediction model; when the first verification result indicates that the prediction error of the initial state prediction model does not exceed an error threshold, the accuracy of the initial state prediction model is verified using a sample test data set to obtain a second verification result; when the second verification result indicates that the accuracy of the initial state prediction model is not less than the accuracy threshold, the initial state prediction model is determined to be the state prediction model.

[0099] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: in the process of iteratively training the support vector machine model using the sample training data set, setting an initial candidate value of the regularization parameter, wherein the regularization parameter represents the degree of penalty for violating the interval, and the violation interval refers to the distance between the sample point in the sample training data set and the decision boundary being less than a predetermined interval; in the process of iteratively cross-validating the initial state prediction model using the sample training data set, updating the initial candidate value using a grid search method until an update termination condition is reached, determining that the current candidate value is the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches a preset model state, and the model state refers to the state of the initial state prediction model in terms of prediction error and accuracy.

[0100] According to another aspect of an embodiment of the present invention, a processor is further provided, and the processor is used to run a program, wherein when the program is run, any one of the above-mentioned transmission line maintenance methods is executed.

[0101] According to another aspect of an embodiment of the present invention, there is further provided a computer program product, comprising computer instructions, which, when executed by a processor, execute any one of the above-mentioned transmission line maintenance methods.

[0102] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0103] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0104] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0105] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0106] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0107] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0108] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for maintaining a power transmission line, characterized in that: include: Receiving operation data of the transmission line at different sampling time points, wherein the operation data is data measured by a plurality of sensors deployed on the transmission line during monitoring of the transmission line, and the plurality of sensors are respectively deployed at different positions on the transmission line; Extracting features from the operating data to obtain feature parameters of the operating data; The characteristic parameters are input into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the characteristic parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each of the multiple sets of training data includes: sample characteristic parameters and sample prediction results corresponding to the sample characteristic parameters; generating a maintenance strategy for the transmission line according to the prediction result; The transmission line is maintained according to the maintenance strategy.

2. The method for maintaining a power transmission line according to claim 1, characterized in that: Receiving operation data of the transmission line at different sampling time points, including at least one of the following: Receive first operation data of the transmission line, wherein the first operation data is data measured by a vibration sensor deployed on the transmission line during monitoring of the transmission line, and the vibration sensor is used to measure the vibration frequency and amplitude of a tower on the transmission line; receiving second operation data of the transmission line, wherein the second operation data is data measured by a temperature sensor deployed on the transmission line during monitoring of the transmission line, and the temperature sensor is used to measure a temperature change value of the transmission line under different environmental conditions; Receive third operating data of the transmission line, wherein the third operating data is data measured by a current sensor deployed on the transmission line during monitoring of the transmission line, and the current sensor is used to measure a current load change value of the transmission line.

3. The method for maintaining a power transmission line according to claim 1, characterized in that: Extracting features from the operating data to obtain feature parameters of the operating data includes: Determine the operation data sampled by the same sensor at different sampling time points as an operation data group; Performing decentralized processing on each of the operation data groups to obtain a target operation data group, wherein the decentralized processing refers to updating each of the operation data in the operation data group to the difference between the operation data and the corresponding data mean, and the data mean is the mean of all the operation data in each of the operation data groups; Sequentially calculating the covariance between each target operation data group and the initial characteristic parameters in all target operation data groups at the same sampling time point to obtain a covariance matrix, wherein the initial characteristic parameter refers to the difference between the operation data and the corresponding data mean; Analyzing the correlation between two initial characteristic parameters in the initial characteristic parameter group according to the covariance matrix to obtain an analysis result, wherein the initial characteristic parameter group is a combination of the initial characteristic parameters in two different target operation data groups at the same sampling time point; Determine, according to the analysis result, the initial feature parameter group whose correlation is not greater than the correlation threshold as the first initial feature parameter group, and determine the initial feature parameter group whose correlation is greater than the correlation threshold as the second initial feature parameter group; Determine two of the initial feature parameters in the first initial feature parameter group and any one of the initial feature parameter groups in the second initial feature parameter group as target feature parameters; The parameter set composed of the target characteristic parameters is reduced in dimension to obtain the characteristic parameters.

4. The method for maintaining a power transmission line according to claim 3, characterized in that: The covariance between each target operation data group and the initial characteristic parameters in all target operation data groups at the same sampling time point is calculated in sequence to obtain a covariance matrix, including: The covariance between each target operation data group and all target operation data groups is calculated in sequence using a first formula, wherein the first formula is: X and Y represent the initial characteristic parameters corresponding to two different sensors, i represents the number of the sampling time point, n represents the total number of the sampling time points, X i and Y i respectively represent the operating data of different sensors at the i-th sampling time point, and Respectively represent the data means corresponding to two different sensors; After obtaining the covariance between each of the target operation data groups and all the target operation data groups, the covariance is arranged in a matrix form to obtain the covariance matrix.

5. The method for maintaining a power transmission line according to claim 3, characterized in that: The correlation between two initial feature parameters in the initial feature parameter group is analyzed according to the covariance matrix to obtain analysis results, including: Calculating the correlation between two initial feature parameters in the initial feature parameter group using a second formula according to the covariance matrix, wherein the second formula is: r(X,Y)=Cov(X,Y) / σXσY, r(X,Y) represents the correlation between the initial feature parameter X and the initial feature parameter Y, Cov(X,Y) represents the covariance between the initial feature parameter X and the initial feature parameter Y, σX represents the first standard deviation of the initial feature parameter X, and σY represents the second standard deviation of the initial feature parameter; Comparing the correlation with the correlation threshold to obtain a comparison result; The comparison result is determined as the analysis result.

6. The method for maintaining a power transmission line according to claim 3, characterized in that: The parameter set composed of the target characteristic parameters is reduced in dimension to obtain the characteristic parameters, including: Determine a set consisting of all the target characteristic parameters as a first characteristic parameter set; Performing eigendecomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors corresponding to the eigenvalues; Sorting the characteristic values ​​in descending order to obtain a sorting result; Selecting a predetermined number of the feature vectors in order according to the sorting result as principal component feature vectors; Constructing a conversion matrix based on the principal component eigenvectors, wherein the conversion matrix is ​​used to perform feature dimension reduction on the first feature parameter set; The first feature parameter set is transformed by using the transformation matrix to obtain a second feature parameter set after feature dimension reduction is performed on the first feature parameter set; Determine the target feature parameter in the second feature parameter set as the feature parameter.

7. The method for maintaining a power transmission line according to claim 1, characterized in that: Before inputting the characteristic parameters into a state prediction model to predict the current state of the transmission line based on the characteristic parameters using the state prediction model to obtain a prediction result, the method further includes: Acquire multiple groups of sample data sets, wherein each of the multiple groups of sample data sets is a data set after feature dimension reduction; Dividing the plurality of groups of the sample data sets to obtain a sample training data set and a sample test data set, wherein the sample training data set is used to train the state prediction model, and the sample test data set is used to test the state prediction model; Iteratively training the support vector machine model using the sample training data set until a first iteration termination condition is reached, determining the support vector machine model obtained by the current training as an initial state prediction model, wherein the first iteration termination condition is at least one of the following: the number of iterations reaches a maximum number of iterations, and the prediction error does not exceed an error threshold; The sample training data set other than the target sample training data set is selected from the sample training data set to train the initial state prediction model, and the target sample training data set is used to iteratively cross-validate the initial state prediction model until a second iteration termination condition is reached to obtain a first verification result, wherein the target sample training data set refers to any one of the multiple groups of sample training data sets, and the second iteration termination condition is that each of the multiple groups of sample training data sets has been used as the target sample training data set to verify the initial state prediction model; When the first verification result indicates that the prediction error of the initial state prediction model does not exceed the error threshold, verifying the accuracy of the initial state prediction model using the sample test data set to obtain a second verification result; When the second verification result indicates that the accuracy of the initial state prediction model is not lower than an accuracy threshold, the initial state prediction model is determined to be the state prediction model.

8. The method for maintaining a power transmission line according to claim 7, characterized in that: Also includes: In the process of iteratively training the support vector machine model using the sample training data set, setting an initial candidate value of a regularization parameter, wherein the regularization parameter represents a degree of penalty for a violation interval, and the violation interval refers to a distance between a sample point in the sample training data set and a decision boundary being less than a predetermined interval; In the process of iterative cross-validation of the initial state prediction model using the sample training data set, the initial candidate value is updated by a grid search method until an update termination condition is reached, and the current candidate value is determined to be the optimal value of the regularization parameter, wherein the update termination condition is: the model state of the initial state prediction model reaches a preset model state, and the model state refers to the state of the initial state prediction model in terms of the prediction error and the accuracy.

9. A maintenance device for a power transmission line, characterized in that: include: A receiving unit, configured to receive operation data of a transmission line at different sampling time points, wherein the operation data is data measured by a plurality of sensors deployed on the transmission line during monitoring of the transmission line, and the plurality of sensors are respectively deployed at different positions on the transmission line; A first acquisition unit, configured to extract features from the operation data to obtain feature parameters of the operation data; A second acquisition unit is used to input the characteristic parameters into a state prediction model, so as to use the state prediction model to predict the line state of the transmission line based on the characteristic parameters to obtain a prediction result, wherein the state prediction model is trained by machine learning using multiple sets of training data, and each of the multiple sets of training data includes: sample characteristic parameters and sample prediction results corresponding to the sample characteristic parameters; A generating unit, configured to generate a maintenance strategy for the transmission line according to the prediction result; A maintenance unit is used to maintain the transmission line according to the maintenance strategy.

10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the power transmission line maintenance method according to any one of claims 1 to 7 is performed.

Images