Power transmission line conductor windage yaw identification method and system

Abstract

The invention discloses a power transmission line conductor windage yaw identification method and system, and the method comprises the steps: arranging a sensor on a power transmission line conductor, and collecting the original operation data of the power transmission line conductor in real time; preprocessing the original operation data to obtain first intermediate operation data; performing space-time position calibration processing on the first intermediate operation data to obtain second intermediate operation data; performing feature extraction on the second intermediate operation data to obtain multi-dimensional features of the second intermediate operation data; performing data fusion on the second intermediate operation data to obtain feature fusion data of the transmission line conductor; and analyzing the feature fusion data of the power transmission line conductor, and judging whether the corresponding power transmission line conductor has a windage yaw effect abnormal condition or not. The power transmission line conductor can be monitored in real time, whether the power transmission line conductor is in a windage yaw abnormal state or not can be quickly and accurately identified, the monitoring efficiency and accuracy are effectively improved, and then the power transmission safety and stability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of windage identification of power transmission lines, and in particular to a method and system for identifying windage of power transmission line conductors. Background Art

[0002] In complex environments, transmission lines face many challenges, especially the impact of wind on overhead transmission lines. Under extreme weather conditions, wind can cause transmission lines to deviate or swing significantly, a phenomenon known as wind deflection. Wind deflection not only increases the risk of conductors coming into contact with insulators and other structures, but can also trigger short circuits, causing power outages or equipment damage, posing a serious threat to the safety of the power system.

[0003] Regular inspections of transmission line conductors are a critical part of ensuring the security and stability of power networks. These inspections are designed to detect and resolve potential problems that could threaten power supply. Traditionally, inspections have relied on human operators who either travel on foot or drive ground vehicles to perform tasks. Although manual inspections are quite effective in identifying obvious damage, this method has certain limitations when it comes to hidden or subtle faults, especially in harsh environmental conditions, and may miss some critical issues and pose safety risks.

[0004] Thanks to the rapid progress of artificial intelligence technology, we have been able to efficiently perform image recognition and monitoring of transmission line equipment using advanced image processing technologies such as convolutional neural networks (CNNs). By using drones to capture towers and wires, collecting image samples, and then using neural networks for rapid analysis, the operating status of transmission lines can be quickly assessed. This method can also safely inspect environments that are too dangerous for manual inspections. Although image perception technology has made significant progress in transmission line monitoring, it still faces a series of challenges in complex natural environments such as dense forests or steep mountainous areas. These challenges include drastic changes in lighting, interference from shadows, and occlusion by vegetation, which increase the difficulty of image processing and target recognition. Despite the use of high-end camera technology, the ability to detect potential hidden dangers may be limited when encountering severe weather conditions or complex terrain, which may reduce the accuracy of inspection results.

[0005] At present, intelligent algorithms mainly identify the status of wires by analyzing the captured images. However, under extreme weather conditions, such as when strong winds cause the wires to deflect, the impact of wind on the wires is significantly enhanced. In this case, whether it is manual inspection or relying on drones to take pictures and transmit data, there are great safety risks, so these methods are not suitable for practical applications. This makes it impossible to monitor whether the wires are deflected in real time, and it is impossible to take preventive measures in time. Usually, we can only observe and repair the wires after they are damaged, which not only leads to a large waste of resources, but also increases the burden of manpower maintenance. Summary of the invention

[0006] The purpose of the present invention is to overcome the shortcomings of the prior art. The present invention provides a method and system for identifying wind deviation of transmission line conductors, which can monitor the transmission line conductors in real time, quickly and accurately identify whether the transmission line conductors are in an abnormal wind deviation state, effectively improve the efficiency and accuracy of monitoring, and thus improve the safety and stability of power transmission.

[0007] The present invention provides a method for identifying windage deviation of a transmission line conductor, the method comprising:

[0008] Arrange sensors on the transmission line conductors, and collect original operation data of the transmission line conductors in real time based on the arranged sensors;

[0009] Preprocessing the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor;

[0010] Performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor;

[0011] Performing feature extraction on the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor;

[0012] Based on the multi-dimensional features, data fusion is performed on the second intermediate operation data of the transmission line conductor to obtain feature fusion data of the transmission line conductor;

[0013] The characteristic fusion data of the transmission line conductor is analyzed to determine whether the corresponding transmission line conductor has abnormal windage effect.

[0014] Furthermore, the step of arranging sensors on the transmission line conductors and collecting the original operation data of the transmission line conductors in real time based on the arranged sensors includes:

[0015] Current sensors are arranged at both ends of the transmission line conductor, and transmission current data of the transmission line conductor is collected based on the current sensors;

[0016] frequency sensors are arranged at both ends of the transmission line conductor and at a plurality of positions between the two ends, and vibration frequency data of the transmission line conductor is collected based on the frequency sensors;

[0017] Arrange a wind speed sensor at a power tower on the transmission line, and collect environmental wind direction data and environmental wind speed data at the power tower of the transmission line based on the wind speed sensor;

[0018] Stress sensors are arranged at both ends of the transmission line conductor and at several positions between the two ends, and displacement data and displacement acceleration data of the transmission line conductor are collected based on the stress sensors.

[0019] Furthermore, the preprocessing of the collected original operation data of the transmission line conductor to obtain the first intermediate operation data of the transmission line conductor includes:

[0020] Process missing values ​​of the collected original operation data of the transmission line conductors;

[0021] Perform noise reduction on the original running data after missing value processing;

[0022] Perform outlier detection on the original running data after noise reduction.

[0023] Furthermore, the missing value processing of the collected original operation data of the transmission line conductor includes:

[0024] Based on the Kalman filter, the position and state of the data point of the next moment of the collected original operation data of the transmission line conductor are predicted through the preset state model, and the predicted value is compared with the actual value;

[0025] When no error occurs in the comparison, the Kalman filter updates the state model based on the position and state of the data point at the next moment of the original operation data;

[0026] When errors occur in the comparison, the Kalman filter performs missing value filling processing based on the predicted value.

[0027] Furthermore, the performing of outlier detection processing on the original operation data after the noise reduction processing includes:

[0028] Based on the Z-score method, outlier detection is performed on the original running data after noise reduction.

[0029] Furthermore, performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor includes:

[0030] Acquire the spatial information and time information corresponding to the first intermediate operation data of the transmission line conductor, perform spatial alignment on the first intermediate operation data based on the spatial information using geographic information system technology or spatial interpolation method, and perform time alignment on the first intermediate operation data based on the time information using timestamp alignment or time series analysis method, and then acquire the second intermediate operation data of the transmission line conductor.

[0031] Furthermore, the extracting features of the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor includes:

[0032] Extracting frequency domain features of the second intermediate operation data based on Fourier transform;

[0033] Extracting time domain features of the second intermediate operation data based on wavelet transform;

[0034] The second intermediate operation data is reduced in dimension based on principal component analysis, and main feature information is extracted.

[0035] Furthermore, the step of fusing the second intermediate operation data of the transmission line conductor based on the multi-dimensional features to obtain the feature fusion data of the transmission line conductor includes:

[0036] Performing data cleaning processing on the second intermediate operation data of the transmission line conductor;

[0037] Normalizing the second intermediate running data after data cleaning;

[0038] Performing time alignment processing and space alignment processing on the normalized second intermediate operation data based on the multidimensional features;

[0039] Based on the unified time dimension and space dimension, data fusion processing is performed on the second intermediate running data after the time alignment processing and the space alignment processing.

[0040] Furthermore, the analyzing the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has an abnormal condition of windage effect includes:

[0041] A wind deviation effect abnormal condition detection model is constructed. After the wind deviation effect abnormal condition detection model is trained, the feature fusion data of the transmission line conductor is analyzed based on the trained wind deviation effect abnormal condition detection model to determine whether the corresponding transmission line conductor has a wind deviation effect abnormal condition, and the detection accuracy is evaluated.

[0042] The present invention also provides a transmission line conductor windage identification system, the system is used to implement the above-mentioned transmission line conductor windage identification method, the system comprises:

[0043] A data acquisition module, wherein the data acquisition module is used to arrange sensors on the transmission line conductors and collect original operation data of the transmission line conductors in real time based on the arranged sensors;

[0044] A data preprocessing module, the data preprocessing module is used to preprocess the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor;

[0045] A time-space position calibration module, the time-space position calibration module is used to perform time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor;

[0046] A feature extraction module, the feature extraction module is used to extract features from the second intermediate operation data of the transmission line conductor, and obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor;

[0047] A data fusion module, the data fusion module is used to fuse the second intermediate operation data of the transmission line conductor based on the multi-dimensional feature to obtain feature fusion data of the transmission line conductor;

[0048] The data analysis module is used to analyze the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has abnormal wind deviation effect.

[0049] The present invention provides a method and system for identifying wind deviation of a transmission line conductor, which adopts a multi-data source method of arranging current sensors, frequency sensors, wind speed sensors and stress sensors to collect original operation data of the transmission line conductor, and then extracts feature information and fuses it to obtain feature fusion data after performing missing value processing, noise reduction processing, outlier detection processing and spatiotemporal position calibration processing on the original operation data, and judges whether an abnormal wind deviation condition occurs based on the feature fusion data. Compared with manual identification or image recognition based on image shooting, the method is less affected by environmental factors and has a smaller judgment error, and can realize real-time monitoring, effectively improving the efficiency and accuracy of monitoring, and thus improving the safety and stability of power transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 is a flow chart of a method for identifying windage deviation of a transmission line conductor in Embodiment 1 of the present invention;

[0052] Figure 2 This is a flow chart of preprocessing the original operation data of the transmission line conductor in the first embodiment of the present invention;

[0053] Figure 3 This is a flow chart of processing missing values ​​of original operation data of a transmission line conductor in the first embodiment of the present invention;

[0054] Figure 4 is a flow chart of performing outlier detection processing on original operation data in Embodiment 1 of the present invention;

[0055] Figure 5 is a flow chart of obtaining multi-dimensional features of second intermediate operation data in Embodiment 1 of the present invention;

[0056] Figure 6 is a flow chart of performing data fusion on the second intermediate operation data in the first embodiment of the present invention;

[0057] Figure 7 It is an architecture diagram of a transmission line conductor wind deviation identification system in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] In the present invention, it should be understood that terms such as "include" or "have" are intended to indicate the existence of features, numbers, steps, behaviors, components, parts or a combination thereof disclosed in the specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, behaviors, components, parts or a combination thereof exist or are added.

[0060] It should also be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0061] Embodiment 1

[0062] Embodiment 1 of the present invention provides a method for identifying wind deviation of a transmission line conductor, the method comprising: arranging sensors on the transmission line conductor, and collecting original operation data of the transmission line conductor in real time based on the arranged sensors; preprocessing the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor; performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor; performing feature extraction on the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor; performing data fusion on the second intermediate operation data of the transmission line conductor based on the multi-dimensional features to obtain feature fusion data of the transmission line conductor; analyzing the feature fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has abnormal wind deviation effect.

[0063] In an optional implementation of this embodiment, Figure 1 As shown, Figure 1 The flowchart of the method for identifying wind deviation of a transmission line conductor in the first embodiment of the present invention is shown, comprising the following steps:

[0064] S101, arranging sensors on the conductors of the transmission line, and collecting original operation data of the conductors of the transmission line in real time based on the arranged sensors;

[0065] In an optional implementation of this embodiment, current sensors are arranged at both ends of the transmission line conductor, and transmission current data of the transmission line conductor is collected based on the current sensors.

[0066] Specifically, a current sensor is arranged at each end of a transmission line conductor that needs to be monitored, and transmission current data of the transmission line conductor is collected based on the arranged current sensors.

[0067] In an optional implementation of this embodiment, frequency sensors are arranged at both ends of the power transmission line conductor and at several positions between the two ends, and vibration frequency data of the power transmission line conductor is collected based on the frequency sensors.

[0068] Specifically, after a frequency sensor is arranged at both ends of the transmission line conductor that needs to be monitored, a frequency sensor is also arranged at a position a certain distance apart between the two ends, and the vibration frequency data of the corresponding position of the transmission line conductor is collected based on the arranged frequency sensors.

[0069] In an optional implementation of this embodiment, a wind speed sensor is arranged at a power tower on the transmission line, and environmental wind direction data and environmental wind speed data at the power tower of the transmission line are collected based on the wind speed sensor.

[0070] Specifically, a wind speed sensor is arranged at a power tower used for assuming a transmission line conductor, and environmental wind direction data and environmental wind speed data at the power tower of the transmission line are collected based on the arranged wind speed sensor.

[0071] In an optional implementation of this embodiment, stress sensors are arranged at both ends of the transmission line conductor and at several positions between the two ends, and displacement data and displacement acceleration data of the transmission line conductor are collected based on the stress sensors.

[0072] Specifically, after arranging a stress sensor at both ends of the transmission line conductor that needs to be monitored, a stress sensor is also arranged at a position a certain distance apart between the two ends, and the displacement data and displacement acceleration data of the transmission line conductor are collected based on the arranged stress sensors.

[0073] In an optional implementation of this embodiment, the acquisition frequency of the above-arranged sensors is 60 Hz-120 Hz.

[0074] It should be noted that since the transmission line conductor is in a wind-deviation state, the vibration frequency and current energy loss of the conductor are different from those in the normal state, and there will be a certain displacement phenomenon. Current sensors, frequency sensors, wind speed sensors and stress sensors are set here to collect transmission current data, vibration frequency data, ambient wind direction data, ambient wind speed data, displacement data and displacement acceleration data, which can be used to accurately analyze and identify whether the transmission line conductor is in a wind-deviation state.

[0075] S102, preprocessing the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor;

[0076] In an optional implementation of this embodiment, missing value processing, noise reduction processing and abnormal value detection processing are performed on the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor.

[0077] In an optional implementation of this embodiment, Figure 2 As shown, Figure 2 A flowchart of preprocessing the original operation data of the transmission line conductor in the first embodiment of the present invention is shown, comprising the following steps:

[0078] S201, performing missing value processing on the collected original operation data of the transmission line conductor;

[0079] In an optional implementation of this embodiment, a filter, median or mode may be used to fill the original operating data, or a decision tree, interpolation method or K-nearest neighbor machine learning algorithm may be used to predict and fill the data. In this embodiment, a Kalman filter is used to predict and fill the data for missing value processing.

[0080] Specifically, Figure 3 As shown, Figure 3 A flowchart of processing missing values ​​of original operation data of a transmission line conductor in Embodiment 1 of the present invention is shown, comprising the following steps:

[0081] S301, predicting the position and state of the data point of the collected original operation data of the transmission line conductor at the next moment through a preset state model based on the Kalman filter, and comparing the predicted value with the actual value;

[0082] In an optional implementation of this embodiment, the Kalman filter can estimate and predict the position and state of the current data at the next moment through a preset state model. In a discrete time system, the system state equation is:

[0083] X(k+1)=F(k)X(k)+G(k)μ(k)+V(k);

[0084] In a discrete time system, the measurement equation is:

[0085] Z(k+1)=H(k+1)X(k+1)+W(k+1);

[0086] Where X(k+1) is the system state at the next moment, F(k) is the first state matrix, X(k) is the system state at the current moment, G(k) is the second state matrix, μ(k) is the control amount of the system at the next moment, V(k) is the prediction process noise, Z(k+1) is the measurement value at the next moment, H(k+1) is the measurement matrix, and W(k+1) is the measurement process noise.

[0087] Its one-step prediction of the system state and covariance is:

[0088]

[0089] P(k+1|k)=F(k)P(k|k)F ′ (k)+Q(k);

[0090] In the formula, is the state prediction result at the next moment, F(k) is the first state matrix, is the optimal state prediction result at the current moment, G(k) is the second state matrix, μ(k) is the control amount of the system at the next moment, and P(k+1|k) is The corresponding covariance, P(k|k) is The corresponding covariance, F′(k) is the transposed matrix of F(k), and Q(k) is the covariance of the system prediction process.

[0091] Furthermore, after the above prediction, a predicted value of the position and state of the data point at the next moment based on the original operation data is obtained, and the predicted value is compared with the actual value of the original operation data at the next moment.

[0092] S302, updating the state model based on the position and state of the data point at the next moment of the original operation data;

[0093] In an optional implementation of this embodiment, when no error occurs in the comparison between the predicted value and the true value at step S301, the Kalman filter updates the state model based on the position and state of the data point at the next moment of the original operation data, and the update formula includes the following:

[0094] The optimal estimate of the system state at time k, that is, the system state update equation is:

[0095]

[0096] In the formula, is the optimal estimate at time k, is the state prediction result at time k, K(k+1) is the Kalman gain, Z(k+1) is the state measurement value at time k, and H(k+1) is the measurement matrix.

[0097] The expression of Kalman gain is:

[0098] K(k+1)=P(k+1|k)H′(k+1) / S(k+1);

[0099] S(k+1)=H(k+1)P(k+1|k)H ' (k+1)+R(k+1);

[0100] Where K(k+1) is the Kalman gain and P(k+1|k) is Corresponding covariance, H′(k+1) is the transposed matrix of H(k+1), S(k+1) is the predicted covariance of the measurement, H(k+1) is the measurement matrix, and R(k+1) is the measurement noise variance.

[0101] Correspondingly, the expression for covariance update is:

[0102] P(k+1|k+1)=[lK(k+1)H(k+1)]P(k+1|k);

[0103] Where P(k+1|k+1) is The corresponding covariance, l is a matrix of 1, K(k+1) is the Kalman gain, H(k+1) is the measurement matrix, and P(k+1|k) is The corresponding covariance.

[0104] S303: Perform missing value filling processing based on the predicted value.

[0105] In an optional implementation of this embodiment, when an error occurs in the comparison between the predicted value and the true value in step S301, the Kalman filter fills in the missing position and state of the data point at the next moment of the original operating data based on the predicted value.

[0106] Furthermore, missing value deletion can also be used here.

[0107] S202, performing noise reduction processing on the original running data after the missing value processing;

[0108] In an optional implementation of this embodiment, the original operating data after the missing value processing is subjected to noise reduction processing by using a filter, including a median filter, a Gaussian filter, or a clustering algorithm.

[0109] S203: Perform outlier detection processing on the original operation data after noise reduction processing.

[0110] In an optional implementation of this embodiment, outlier detection processing is performed on the original operating data after noise reduction processing based on the Z-score method.

[0111] Specifically, Figure 4 As shown, Figure 4 The flowchart of performing outlier detection processing on original operation data in the first embodiment of the present invention is shown, including the following steps:

[0112] S401, preparing the original operation data;

[0113] In an optional implementation of this embodiment, before performing outlier detection processing, it is necessary to ensure that the original operating data is numerical data and that there are no NaN values. If there are any, filling or deletion processing is required.

[0114] S402, calculating the mean and standard deviation of the original operation data;

[0115] In an optional implementation of this embodiment, the mean is the average value of all data points in the original operation data set, and the standard deviation is the deviation of each data point in the original operation data set from the mean.

[0116] S403, calculating the Z-score value of the original running data;

[0117] In an optional implementation of this embodiment, a Z-score formula is applied to each data point in the original operation data set for calculation, and the formula is as follows:

[0118]

[0119] In the formula, Z i is the Z-score value of the ith data point, μ is the mean, and σ is the standard deviation.

[0120] Furthermore, the Z-score values ​​corresponding to each data point are combined into a new Z-score value list.

[0121] S404, identifying abnormal values ​​based on the Z-score value of the original running data;

[0122] In an optional implementation of this embodiment, under normal circumstances, the Z-score value of a data point between [-2, 2] or [-3, 3] is considered to be a normal value, so the threshold is set to 2 or 3 here. When the absolute value of the Z-score value of a data point is greater than 2 or greater than 3, the data point is considered to be an outlier and is marked.

[0123] S405: Process the abnormal values ​​of the original operation data.

[0124] In an optional implementation of this embodiment, there are three ways to process the outliers of the original operating data. The first is to delete the outliers. If the outliers have a significant impact on the data set, consider deleting the data points corresponding to the outliers. The second is to replace the outliers by other filling values ​​such as the mean, median, mode, etc. The third is to correct the outliers and make reasonable corrections based on conventional data comparisons.

[0125] S103, performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor;

[0126] In an optional implementation of the present embodiment, spatial information and time information corresponding to the first intermediate operation data of the transmission line conductor are obtained, the first intermediate operation data are spatially aligned based on the spatial information using geographic information system technology or spatial interpolation method, and the first intermediate operation data are time-aligned based on the time information using timestamp alignment or time series analysis method, and then the second intermediate operation data of the transmission line conductor is obtained.

[0127] In an optional implementation of the present embodiment, the use of geographic information system technology to spatially align the first intermediate operation data based on the spatial information includes: obtaining the geographic location information of all sensors based on the geographic information system, uniformly expressing it in the form of longitude and latitude or other coordinate systems, and using practical GIS tools (ArcGIS, QGIS) or geographic libraries (Python's pyproj library) to perform coordinate conversion.

[0128] In an optional implementation of this embodiment, the spatial interpolation method is used to perform spatial alignment on the first intermediate operation data based on the spatial information, including: obtaining the geographic location information of all sensors, uniformly expressing it in the form of longitude and latitude or other coordinate systems, and since the spatial interpolation method is implemented by inverse distance weighted interpolation (IDW), assuming that close points have a greater impact on a certain position, the inverse distance weighted method performs weighted interpolation on known points according to the distance, and the specific formula is:

[0129]

[0130] In the formula, Z(x) is the data point that needs spatial interpolation, Z(x i ) is the spatial interpolation of known data points, d is the distance between two points, and p is the weight index.

[0131] In an optional implementation of this embodiment, the adopting a time series analysis method to perform time alignment on the first intermediate operation data based on the time information includes:

[0132] When the sampling frequencies of sensors are inconsistent, data with different sampling frequencies can be aligned by linear interpolation. Therefore, according to the difference in timestamps, data with low sampling rate can be interpolated to the time point of data with high sampling rate.

[0133] Specifically, assume that the time series is {t1, t2, …, t n}, the first intermediate running data is {x1,x2,……,x n}, then the linear interpolation formula includes:

[0134]

[0135] Where l(n) is linear interpolation.

[0136] Furthermore, it is necessary to evaluate the time consistency after calibration by comparing the time differences of key times before and after calibration (such as sensor failure, vibration response time and other factors); in addition, the time consistency after calibration can be evaluated by calculating the root mean square error (RMSE) and mean absolute error (MAE) of the time difference before and after calibration.

[0137] S104, performing feature extraction on the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor;

[0138] In an optional implementation of this embodiment, Figure 5 As shown, Figure 5 A flow chart of obtaining multi-dimensional features of second intermediate operation data in the first embodiment of the present invention is shown, including the following steps:

[0139] S501, extracting frequency domain features of the second intermediate operation data based on Fourier transform;

[0140] In an optional implementation of this embodiment, before Fourier transform is performed, the second intermediate running data needs to be preprocessed, including de-averaging, de-noising and sampling processing.

[0141] In an optional implementation of this embodiment, the frequency domain features are extracted by discrete Fourier transform, and the formula includes:

[0142]

[0143] Where X(k) is the frequency domain representation of the data, representing the kth frequency component, x(n) is a discrete time domain signal with a length of n, k is the index of the frequency component, and j is the imaginary unit.

[0144] Furthermore, for frequency domain feature analysis, the amplitude spectrum (the amplitude of the signal at different frequencies, used to indicate the strength of the signal at each frequency component) and the phase spectrum (indicating the phase information of the signal at each frequency) are used for analysis. The specific extraction parameters include:

[0145] Main frequency: The frequency where the maximum frequency component of the signal is located. The frequency component with the largest amplitude is found through the amplitude spectrum. The formula includes:

[0146] f peak = argmax(|X(k)|);

[0147] RMS frequency: The RMS value of the frequency, which reflects the degree of balance of the signal frequency distribution. The formula includes:

[0148]

[0149] Spectral entropy: used to measure the flatness of the spectrum and reflect the frequency complexity of the signal. The formula includes:

[0150]

[0151] S502, extracting time domain features of the second intermediate operation data based on wavelet transform;

[0152] In an optional implementation of this embodiment, before performing wavelet transform, the second intermediate running data needs to be preprocessed, including denoising and standardization or normalization processing.

[0153] In an optional implementation of this embodiment, it is first necessary to select the number of wavelet transform decomposition layers. The wavelet transform divides the signal into low-frequency components (approximation coefficients) and high-frequency components (detail coefficients) by recursive decomposition. The number of decomposition layers is related to the sampling frequency and signal duration of the signal. Generally, it is necessary to select the optimal number of decomposition layers based on experience or cross-validation.

[0154] Secondly, wavelet transform decomposition is performed on the second intermediate running data to decompose the original signal into detail coefficients (high frequency part) and approximation coefficients (low frequency part) of different frequency bands (levels).

[0155] Next, wavelet decomposition coefficients (detail coefficients and approximation coefficients) at different levels are extracted, which can describe the local time domain and frequency domain characteristics of the signal.

[0156] Finally, the feature vector is constructed by calculating the statistical characteristics of the wavelet coefficients of each layer, such as the mean, standard deviation, and energy.

[0157] S503: Perform dimension reduction on the second intermediate operation data based on principal component analysis, and extract main feature information.

[0158] In an optional implementation of this embodiment, before performing the principal component analysis process, the second intermediate running data needs to be preprocessed, including taking the mean and standardization or normalization processing.

[0159] In an optional implementation of this embodiment, it is first necessary to calculate the covariance matrix, which is calculated for the preprocessed data set to measure the correlation between different features of the data. The formula includes:

[0160]

[0161] Where C is the covariance matrix, X i is the i-th sample data, μ is the mean.

[0162] Secondly, calculate the eigenvalues ​​and eigenvectors, perform eigenvalue decomposition on the covariance matrix, and obtain the eigenvalues ​​and corresponding eigenvectors. The eigenvalues ​​represent the variance of each principal component, and the eigenvectors represent the directions of these principal components.

[0163] Next, select the principal components. According to the size of the eigenvalues, select the eigenvectors corresponding to the first k largest eigenvalues ​​as the principal components. The number of principal components k selected can be determined by the cumulative variance contribution rate. Usually, the principal components that can explain more than 90% of the variance are selected. The cumulative variance contribution rate formula includes:

[0164]

[0165] In the formula, λ i is the i-th eigenvalue.

[0166] Finally, the original data is projected into the low-dimensional space composed of the selected principal components to generate a new low-dimensional feature vector.

[0167] Y=X·W

[0168] In the formula, X is the original matrix, W is the matrix composed of selected eigenvectors, and Y is the data after dimensionality reduction.

[0169] S105, performing data fusion on the second intermediate operation data of the transmission line conductor based on the multi-dimensional features to obtain feature fusion data of the transmission line conductor;

[0170] In an optional implementation of this embodiment, Figure 6 As shown, Figure 6 A flow chart of performing data fusion on the second intermediate operation data in the first embodiment of the present invention is shown, including the following steps:

[0171] S601, performing data cleaning processing on the second intermediate operation data of the transmission line conductor;

[0172] In an optional implementation of this embodiment, the second intermediate operation data of the transmission line conductor is processed to remove duplicate data, remove missing values, and correct data errors.

[0173] S602, normalizing the second intermediate operation data after data cleaning;

[0174] In an optional implementation of this embodiment, normalization processing is performed here to ensure that the data are analyzed and pattern recognized at the same scale.

[0175] S603, performing time alignment processing and space alignment processing on the normalized second intermediate operation data based on the multidimensional features;

[0176] In an optional implementation of this embodiment, time alignment processing and space alignment processing are performed on the normalized second intermediate running data based on the multidimensional features.

[0177] S604: Perform data fusion processing on the second intermediate operation data after the time alignment processing and the space alignment processing based on the unified time dimension and space dimension.

[0178] In an optional implementation of the present embodiment, data fusion processing is performed on the second intermediate running data after the time alignment processing and the space alignment processing on the unified time dimension and space dimension based on a preset data fusion standard.

[0179] S106: Analyze the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has abnormal windage effect.

[0180] In an optional implementation of the present embodiment, a wind deviation effect abnormal condition detection model is constructed. After the wind deviation effect abnormal condition detection model is trained, the feature fusion data of the transmission line conductor is analyzed based on the trained wind deviation effect abnormal condition detection model to determine whether the corresponding transmission line conductor has a wind deviation effect abnormal condition, and the detection accuracy is evaluated.

[0181] What's more, when an abnormal situation occurs, the early warning mechanism is automatically triggered, and possible causes of the failure and recommended solutions are provided.

[0182] In summary, Embodiment 1 of the present invention provides a method for identifying wind deviation of a transmission line conductor, which sets a multi-data source method of current sensors, frequency sensors, wind speed sensors and stress sensors to collect original operating data of the transmission line conductor, and then performs missing value processing, noise reduction processing, outlier detection processing, and spatiotemporal position calibration processing on the original operating data, extracts feature information and fuses it to obtain feature fusion data, and judges whether an abnormal wind deviation condition occurs based on the feature fusion data. Compared with manual identification or image recognition based on image shooting, it is less affected by environmental factors and has a smaller judgment error, and can realize real-time monitoring, effectively improving the efficiency and accuracy of monitoring, thereby improving the safety and stability of power transmission.

[0183] Embodiment 2

[0184] Embodiment 2 of the present invention provides a transmission line conductor wind deviation identification system, which is used to implement the transmission line conductor wind deviation identification method in embodiment 1, and the system includes a data acquisition module, a data preprocessing module, a time-space position calibration module, a feature extraction module, a data fusion module, and a data analysis module.

[0185] In an optional implementation of this embodiment, Figure 7 As shown, Figure 7 The following is an architecture diagram of a transmission line conductor wind deviation identification system in Embodiment 2 of the present invention, wherein the system comprises:

[0186] A data acquisition module 10, wherein the data acquisition module 10 is used to arrange sensors on the transmission line conductors, and collect original operation data of the transmission line conductors in real time based on the arranged sensors;

[0187] A data preprocessing module 20, wherein the data preprocessing module 20 is used to preprocess the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor;

[0188] A time-space position calibration module 30, the time-space position calibration module 30 is used to perform time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor;

[0189] A feature extraction module 40, the feature extraction module 40 is used to extract features from the second intermediate operation data of the transmission line conductor, and obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor;

[0190] A data fusion module 50, the data fusion module 50 is used to fuse the second intermediate operation data of the transmission line conductor based on the multi-dimensional feature to obtain feature fusion data of the transmission line conductor;

[0191] The data analysis module 60 is used to analyze the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has an abnormal condition of wind deviation effect.

[0192] In summary, embodiment 2 of the present invention provides a transmission line conductor wind deviation identification system, which is used to implement the transmission line conductor wind deviation identification method in embodiment 1, and sets a multi-data source method of current sensors, frequency sensors, wind speed sensors and stress sensors to collect the original operation data of the transmission line conductor, and then performs missing value processing, noise reduction processing, outlier detection processing, and spatiotemporal position calibration processing on the original operation data, extracts feature information and fuses it to obtain feature fusion data, and judges whether an abnormal wind deviation condition occurs based on the feature fusion data. Compared with manual identification or image recognition based on image shooting, it is less affected by environmental factors and has a smaller judgment error, and can realize real-time monitoring, effectively improving the efficiency and accuracy of monitoring, and thus improving the safety and stability of power transmission.

[0193] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0194] In addition, the embodiments of the present invention are described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for identifying wind deviation of a transmission line conductor, characterized in that: The method comprises: Arrange sensors on the transmission line conductors, and collect original operation data of the transmission line conductors in real time based on the arranged sensors; Preprocessing the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor; Performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor; Performing feature extraction on the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor; Based on the multi-dimensional features, data fusion is performed on the second intermediate operation data of the transmission line conductor to obtain feature fusion data of the transmission line conductor; The characteristic fusion data of the transmission line conductor is analyzed to determine whether the corresponding transmission line conductor has abnormal windage effect.

2. The method for identifying windage deviation of a power transmission line conductor according to claim 1, characterized in that: The step of arranging sensors on the transmission line conductors and collecting original operation data of the transmission line conductors in real time based on the arranged sensors includes: Current sensors are arranged at both ends of the transmission line conductor, and transmission current data of the transmission line conductor is collected based on the current sensors; frequency sensors are arranged at both ends of the transmission line conductor and at a plurality of positions between the two ends, and vibration frequency data of the transmission line conductor is collected based on the frequency sensors; Arrange a wind speed sensor at a power tower on the transmission line, and collect environmental wind direction data and environmental wind speed data at the power tower of the transmission line based on the wind speed sensor; Stress sensors are arranged at both ends of the transmission line conductor and at several positions between the two ends, and displacement data and displacement acceleration data of the transmission line conductor are collected based on the stress sensors.

3. The method for identifying windage deviation of a transmission line conductor according to claim 1, characterized in that: The preprocessing of the collected original operation data of the transmission line conductor to obtain the first intermediate operation data of the transmission line conductor comprises: Process missing values ​​of the collected original operation data of the transmission line conductors; Perform noise reduction on the original running data after missing value processing; Perform outlier detection on the original running data after noise reduction.

4. The method for identifying wind deviation of a transmission line conductor according to claim 3, characterized in that: The missing value processing of the collected original operation data of the transmission line conductor includes: Based on the Kalman filter, the position and state of the data point of the next moment of the collected original operation data of the transmission line conductor are predicted through the preset state model, and the predicted value is compared with the actual value; When no error occurs in the comparison, the Kalman filter updates the state model based on the position and state of the data point at the next moment of the original operation data; When errors occur in the comparison, the Kalman filter performs missing value filling processing based on the predicted value.

5. The method for identifying windage deviation of a power transmission line conductor according to claim 3, characterized in that: The performing outlier detection processing on the original operation data after the noise reduction processing comprises: Based on the Z-score method, outlier detection is performed on the original running data after noise reduction.

6. The method for identifying windage deviation of a power transmission line conductor according to claim 1, characterized in that: The performing time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain the second intermediate operation data of the transmission line conductor comprises: Acquire the spatial information and time information corresponding to the first intermediate operation data of the transmission line conductor, perform spatial alignment on the first intermediate operation data based on the spatial information using geographic information system technology or spatial interpolation method, and perform time alignment on the first intermediate operation data based on the time information using timestamp alignment or time series analysis method, and then acquire the second intermediate operation data of the transmission line conductor.

7. The method for identifying windage deviation of a power transmission line conductor according to claim 1, characterized in that: The extracting features of the second intermediate operation data of the transmission line conductor to obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor comprises: Extracting frequency domain features of the second intermediate operation data based on Fourier transform; Extracting time domain features of the second intermediate operation data based on wavelet transform; The second intermediate operation data is reduced in dimension based on principal component analysis, and main feature information is extracted.

8. The method for identifying windage deviation of a power transmission line conductor according to claim 1, characterized in that: The step of fusing the second intermediate operation data of the transmission line conductor based on the multi-dimensional features to obtain the feature fusion data of the transmission line conductor comprises: Performing data cleaning processing on the second intermediate operation data of the transmission line conductor; Normalizing the second intermediate running data after data cleaning; Performing time alignment processing and space alignment processing on the normalized second intermediate operation data based on the multidimensional features; Based on the unified time dimension and space dimension, data fusion processing is performed on the second intermediate running data after the time alignment processing and the space alignment processing.

9. The method for identifying windage deviation of a power transmission line conductor according to claim 1, characterized in that: The analyzing the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has an abnormal condition of windage effect includes: A wind deviation effect abnormal condition detection model is constructed. After the wind deviation effect abnormal condition detection model is trained, the feature fusion data of the transmission line conductor is analyzed based on the trained wind deviation effect abnormal condition detection model to determine whether the corresponding transmission line conductor has a wind deviation effect abnormal condition, and the detection accuracy is evaluated.

10. A transmission line conductor wind deviation identification system, characterized in that: The system is used to implement the method for identifying wind deviation of a transmission line conductor according to any one of claims 1 to 9, and the system comprises: A data acquisition module, wherein the data acquisition module is used to arrange sensors on the transmission line conductors and collect original operation data of the transmission line conductors in real time based on the arranged sensors; A data preprocessing module, the data preprocessing module is used to preprocess the collected original operation data of the transmission line conductor to obtain first intermediate operation data of the transmission line conductor; A time-space position calibration module, the time-space position calibration module is used to perform time-space position calibration processing on the first intermediate operation data of the transmission line conductor to obtain second intermediate operation data of the transmission line conductor; A feature extraction module, the feature extraction module is used to extract features from the second intermediate operation data of the transmission line conductor, and obtain multi-dimensional features of the second intermediate operation data of the transmission line conductor; A data fusion module, the data fusion module is used to fuse the second intermediate operation data of the transmission line conductor based on the multi-dimensional feature to obtain feature fusion data of the transmission line conductor; The data analysis module is used to analyze the characteristic fusion data of the transmission line conductor to determine whether the corresponding transmission line conductor has abnormal wind deviation effect.

Images