Cross arm capacitive stress data precision optimization method and system

Abstract

The application discloses a cross arm capacitive stress data precision optimization method and system, relates to the technical field of power transmission line mechanical monitoring, and comprises the following steps: collecting cross arm capacitive change signals through a capacitive sensor and performing temperature compensation, establishing a multi-measuring-point mutual checking mechanism based on the compensated data, and constructing a measuring-point interval constraint relationship calculation theory capacitance value by using a cross arm mechanical balance condition. After the deviation between the actual and theoretical capacitance values is taken as a correction amount and is compensated, a segmented conversion strategy is adopted to obtain stress data and reconstruct the stress data. By calculating the residual error of the reconstructed data and the mutual checking data, a spatial gradient is extracted to identify a sudden change measuring point and is corrected, and finally, optimized stress data are obtained, so that cross arm bearing state evaluation and abnormal early warning are realized. The method improves stress monitoring precision and has important engineering application value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of transmission line mechanical monitoring, in particular to a cross arm capacitive stress data precision optimization method and system. BACKGROUND

[0002] With the continuous expansion of the power system, the cross arm of the transmission line as an important load-bearing component, its stress monitoring is of great significance to ensure the safe operation of the power grid. Capacitive sensors are widely used in cross arm stress monitoring due to their high sensitivity, strong anti-interference ability and other advantages.

[0003] The cross arm stress monitoring method mainly relies on single-point capacitive data to directly convert the stress value. This method is easily affected by temperature changes, measurement errors and other factors, resulting in insufficient stress data precision.

[0004] The current stress data processing method often uses a unified conversion coefficient, which fails to process different regions of the cross arm according to their stress characteristics, resulting in a deviation between the conversion result and the actual stress state. The existing method lacks an effective identification and correction mechanism for abnormal data, affecting the accuracy of the stress monitoring result. SUMMARY

[0005] The purpose of the present application is to provide a cross arm capacitive stress data precision optimization method and system, which aims to at least solve one of the technical problems existing in the prior art.

[0006] The technical solution of the present application is: a cross arm capacitive stress data precision optimization method, comprising the following steps:

[0007] Collecting the capacitive change signal of the cross arm by the capacitive sensor to obtain the original capacitive data sequence, and performing temperature compensation processing on the original capacitive data sequence to obtain the temperature-compensated capacitive data sequence;

[0008] Based on the temperature-compensated capacitive data sequence, a multi-measurement-point mutual verification mechanism is established, multiple measurement points on the cross arm are selected, and the constraint relationship between the capacitive data of each measurement point is constructed according to the mechanical equilibrium condition of the cross arm to calculate the theoretical capacitive value of each measurement point;

[0009] The deviation between the actual capacitive value and the theoretical capacitive value is used as a correction amount to compensate the data of each measurement point to obtain the mutual verification capacitive data sequence;

[0010] The mutual verification capacitive data sequence is segmented and converted into stress values to obtain a preliminary stress data sequence, and the preliminary stress data sequence is converted back into capacitive data to obtain a reconstructed capacitive data sequence;

[0011] A residual of the reconstructed capacitance data sequence and the mutual checked capacitance data sequence is calculated, a spatial gradient of the residual along a position of a measuring point is extracted, a sudden change measuring point is identified, a conversion relationship of adjacent measuring points is used to correct the sudden change measuring point, and an optimized cross arm stress data sequence is obtained;

[0012] A bearing state of the cross arm is evaluated based on the optimized cross arm stress data, a stress abnormal area is identified, and early warning information is generated.

[0013] A capacitance change signal of the cross arm is collected by a capacitive sensor to obtain an original capacitance data sequence, and a temperature compensation process is performed on the original capacitance data sequence to obtain a temperature compensated capacitance data sequence, including:

[0014] A capacitance change signal of a preset measuring point on the surface of the cross arm is collected by a capacitive sensor, and temperature data of the preset measuring point position is synchronously collected to form an original capacitance data sequence containing the capacitance change signal and the temperature data;

[0015] A temperature distribution difference value between the preset measuring points is calculated based on the temperature data, and a temperature gradient change amount of the preset measuring points is determined according to the temperature distribution difference value;

[0016] A temperature compensation coefficient is generated according to a mapping relationship between the temperature gradient change amount and the capacitance change signal, and a temperature compensation correction matrix is constructed according to a positional relationship of the preset measuring points;

[0017] Temperature change trends of the preset measuring points in the horizontal and vertical directions are calculated respectively, and a bidirectional compensation factor is constructed by combining the temperature change trends and the temperature compensation correction matrix;

[0018] The capacitance change signal in the original capacitance data sequence is compensated layer by layer according to the bidirectional compensation factor to obtain the temperature compensated capacitance data sequence.

[0019] A multi-measuring point mutual checking mechanism is established based on the temperature compensated capacitance data sequence, a plurality of measuring points on the cross arm are selected, a constraint relationship between capacitance data of each measuring point is constructed according to a mechanical balance condition of the cross arm, and a theoretical capacitance value of each measuring point is calculated, including:

[0020] A trend analysis is performed on the temperature compensated capacitance data sequence, a capacitance value change trend between adjacent measuring points on the cross arm is calculated, a capacitance change amplitude and a change direction between the measuring points are extracted, a capacitance change correlation degree value between the measuring points is generated, and a capacitance change correlation matrix is constructed;

[0021] A capacitance change law between the measuring points is analyzed based on the capacitance change correlation matrix, a mutual influence degree between the measuring points is extracted, a measuring point capacitance data checking criterion is established, and a multi-measuring point mutual checking mechanism is formed;

[0022] According to the multi-point mutual checking mechanism, multiple measuring points on the cross arm are selected to meet the stress transmission requirements, and the distribution positions of the measuring points are determined;

[0023] According to the mechanical balance condition of the cross arm and the distribution positions of the measuring points, the stress transmission relationship between the multiple measuring points is calculated, and a constraint relationship between the multiple measuring point capacitance data is constructed;

[0024] The distribution positions of the multiple measuring points are substituted into the constraint relationship between the multiple measuring point capacitance data to construct a constraint equation group, the constraint equation group is solved, and the theoretical capacitance values of the measuring points are calculated.

[0025] The deviation between the actual capacitance value and the theoretical capacitance value is used as a correction amount to compensate the data of each measuring point, and a mutual checking capacitance data sequence is obtained, including:

[0026] The actual capacitance values of the measuring points are obtained, the difference sequence between the actual capacitance values and the theoretical capacitance values is calculated, and a measuring point capacitance deviation sequence is generated according to the difference sequence;

[0027] According to the stress transmission path determined according to the stress characteristics of the cross arm, the variation characteristics of the measuring point capacitance deviation sequence are analyzed along the stress transmission path, the deviation variation amount and the transmission law between adjacent measuring points are extracted, and a measuring point compensation parameter is constructed;

[0028] The spatial distribution of the deviation variation amount is corrected by using the measuring point compensation parameter, and a dynamic compensation coefficient is generated based on the transmission law;

[0029] The dynamic compensation coefficient is used to modify and calculate the actual capacitance measurement value of each measuring point in sequence according to the stress transmission path, and a mutual checking capacitance data sequence is obtained.

[0030] The mutual checking capacitance data sequence is segmented and converted into a stress value to obtain a preliminary stress data sequence, and the preliminary stress data sequence is converted back into capacitance data to obtain a reconstructed capacitance data sequence, including:

[0031] The capacitance variation trend of the measuring points in the mutual checking capacitance data sequence is analyzed, the capacitance variation amplitude between adjacent measuring points is calculated, the mutation position is identified according to the capacitance variation amplitude, and the data segmentation interval is determined;

[0032] The stress deformation characteristics of the measuring points in the data segmentation interval are extracted, the corresponding relationship between the stress deformation characteristics and the capacitance variation is analyzed, the capacitance-stress conversion coefficient of each data segmentation interval is generated, and a segmented mapping relationship from capacitance to stress is constructed;

[0033] The segmented mapping relationship is used to convert and calculate the mutual checking capacitance data sequence, and the numerical value is smoothly connected through the boundary of the segmentation to obtain a preliminary stress data sequence;

[0034] The capacitor-stress conversion coefficient is reversely configured to establish a reverse calculation relationship from stress to capacitor, the preliminary stress data sequence is reversely converted to obtain a reconstructed capacitor data sequence.

[0035] Residuals of the reconstructed capacitor data sequence and the mutually checked capacitor data sequence are calculated, spatial gradients of the residuals along the positions of the measuring points are extracted, abrupt measuring points are identified, the abrupt measuring points are corrected by using conversion relationships of adjacent measuring points to obtain an optimized cross arm stress data sequence including:

[0036] Numerical differences of the reconstructed capacitor data sequence and the mutually checked capacitor data sequence at the positions of the measuring points are calculated to generate measuring point residual data, and residual variation amounts between the measuring points are calculated according to the measuring point residual data.

[0037] Residual distribution functions are constructed by using the measuring point residual data and the residual variation amounts, spatial direction derivatives of the residual distribution functions at the positions of the measuring points are calculated to obtain residual spatial gradient values.

[0038] Abrupt measuring points are determined by analyzing the measuring point positions whose residual spatial gradient values exceed a preset gradient range, and capacitor variation amounts of the measuring points on both sides of the abrupt measuring points are extracted to generate a measuring point conversion function.

[0039] The measuring point conversion function is applied to the stress data at the abrupt measuring points to correct and compensate to obtain the optimized cross arm stress data sequence.

[0040] A bearing state of the cross arm is evaluated based on the optimized cross arm stress data, stress abnormal regions are identified, and warning information is generated including:

[0041] Stress distribution characteristics of the measuring points in the optimized cross arm stress data sequence are analyzed, stress variation trends of the measuring points are extracted, and a cross arm stress distribution map is constructed.

[0042] Stress concentration degrees of regions of the cross arm are calculated according to the cross arm stress distribution map, a stress state of the cross arm is evaluated, and a cross arm bearing state evaluation result is generated.

[0043] Stress over-limit regions are identified based on the cross arm bearing state evaluation result, stress aggregation degrees in the stress over-limit regions are calculated, and a stress risk evaluation index is constructed.

[0044] The stress risk evaluation index and the stress aggregation degree are combined to make a hierarchical judgment, and warning information including a risk level and a region location is generated.

[0045] An embodiment of the present application provides a cross arm capacitor type stress data precision optimization system, the system including:

[0046] The data acquisition and temperature compensation unit is configured to acquire a capacitance change signal of the cross arm through a capacitive sensor, obtain an original capacitance data sequence, and perform temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence.

[0047] The multi-measurement-point mutual verification unit is configured to establish a multi-measurement-point mutual verification mechanism based on the temperature-compensated capacitance data sequence, select multiple measurement points on the cross arm, construct a constraint relationship between the capacitance data of each measurement point according to a mechanical balance condition of the cross arm, and calculate a theoretical capacitance value of each measurement point.

[0048] The data compensation unit is configured to compensate the measurement point data by taking a deviation between the actual capacitance value and the theoretical capacitance value as a correction amount to obtain a mutual-verification capacitance data sequence.

[0049] The data conversion unit is configured to convert the mutual-verification capacitance data sequence into stress values in segments to obtain a preliminary stress data sequence, and convert the preliminary stress data sequence back into capacitance data to obtain a reconstructed capacitance data sequence.

[0050] The abrupt measurement point correction unit is configured to calculate a residual error between the reconstructed capacitance data sequence and the mutual-verification capacitance data sequence, extract a spatial gradient of the residual error along the measurement point positions, identify an abrupt measurement point, correct the abrupt measurement point by using a conversion relationship of adjacent measurement points, and obtain an optimized cross arm stress data sequence.

[0051] The early warning unit is configured to evaluate a bearing state of the cross arm based on the optimized cross arm stress data, identify a stress abnormal area, and generate early warning information.

[0052] The present application has the following advantages:

[0053] The present application establishes a multi-measurement-point mutual verification mechanism, combines temperature compensation processing and a segmented conversion strategy, and significantly improves the accuracy of cross arm stress monitoring data. The residual error spatial gradient analysis method can accurately identify an abrupt measurement point and effectively correct the abrupt measurement point by using a conversion relationship of adjacent measurement points, thereby ensuring the reliability of the stress data. The bearing state evaluation based on the optimized stress data enables timely discovery and early warning of a stress abnormal area. The present application not only solves the problem that the traditional single-point monitoring method is easily disturbed by external factors, but also overcomes the accuracy deviation caused by a unified conversion coefficient, thereby providing more reliable data support for the safe operation monitoring of the cross arm and having important engineering application value. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 A flowchart of a cross arm capacitive stress data accuracy optimization method according to an embodiment of the present application;

[0055] Figure 2 A cross arm stress data abnormal value detection and correction flowchart according to an embodiment of the present application. DETAILED DESCRIPTION

[0056] As Figure 1 shown, Figure 1 A flow chart of a cross arm capacitive stress data precision optimization method provided by an embodiment of the application, the method comprising the following steps:

[0057] Collecting a capacitive change signal of the cross arm through a capacitive sensor to obtain an original capacitive data sequence, and performing temperature compensation processing on the original capacitive data sequence to obtain a temperature-compensated capacitive data sequence;

[0058] Based on the temperature-compensated capacitive data sequence, a multi-measurement-point mutual verification mechanism is established, multiple measurement points on the cross arm are selected, a constraint relationship between the capacitive data of each measurement point is constructed according to the mechanical equilibrium condition of the cross arm, and the theoretical capacitive value of each measurement point is calculated;

[0059] The deviation of the actual capacitive value from the theoretical capacitive value is taken as a correction amount to compensate the data of each measurement point, and a mutual-verified capacitive data sequence is obtained;

[0060] The mutual-verified capacitive data sequence is segmented and converted into stress values to obtain a preliminary stress data sequence, and the preliminary stress data sequence is converted back into capacitive data to obtain a reconstructed capacitive data sequence;

[0061] The residual error between the reconstructed capacitive data sequence and the mutual-verified capacitive data sequence is calculated, the spatial gradient of the residual error along the measurement point position is extracted, the abrupt measurement points are identified, the conversion relationship of adjacent measurement points is used to correct the abrupt measurement points, and an optimized cross arm stress data sequence is obtained;

[0062] Based on the optimized cross arm stress data, the bearing state of the cross arm is evaluated, the stress abnormal area is identified, and warning information is generated.

[0063] Collecting a capacitive change signal of the cross arm through a capacitive sensor to obtain an original capacitive data sequence, and performing temperature compensation processing on the original capacitive data sequence to obtain a temperature-compensated capacitive data sequence includes:

[0064] Collecting a capacitive change signal of a preset measurement point on the surface of the cross arm through a capacitive sensor, and synchronously collecting temperature data of the preset measurement point position to form an original capacitive data sequence containing the capacitive change signal and the temperature data;

[0065] Based on the temperature data, a temperature distribution difference value between the preset measurement points is calculated, and a temperature gradient change amount of the preset measurement points is determined according to the temperature distribution difference value;

[0066] A temperature compensation coefficient is generated according to the mapping relationship between the temperature gradient change amount and the capacitive change signal, and a temperature compensation correction matrix is constructed according to the positional relationship of the preset measurement points.

[0067] The temperature variation trends of the preset measuring points in the lateral and longitudinal directions are calculated respectively, and the temperature variation trends are combined with a temperature compensation correction matrix to construct a bidirectional compensation factor;

[0068] The capacitance variation signal in the original capacitance data sequence is compensated layer by layer according to the bidirectional compensation factor, to obtain a temperature-compensated capacitance data sequence.

[0069] The capacitive sensor is installed on the surface of the cross arm for collecting the capacitance variation signal of the cross arm. The sensor adopts a grid arrangement to form a plurality of preset measuring points on the surface of the cross arm. Each preset measuring point is equipped with a temperature sensor to realize synchronous collection of the capacitance variation signal and temperature data. The sensitivity of the capacitive sensor is set to 0.01 pF, and the sampling frequency is 100 Hz. The accuracy of the temperature sensor is ±0.1℃, and the sampling frequency is consistent with that of the capacitive sensor. The collected data includes the capacitance variation signal value and the corresponding temperature value, which constitutes the original capacitance data sequence and is stored in the data buffer.

[0070] The original capacitance data sequence is preprocessed to filter out high-frequency noise and outliers. The preprocessing adopts a median filtering method, and the filtering window size is 5 sampling points. The outlier judgment standard is 3 times the standard deviation of the average value of the previous and subsequent 10 sampling points. The preprocessed data is used for subsequent temperature compensation processing.

[0071] The temperature compensation processing first calculates the temperature distribution difference values between the preset measuring points. For adjacent measuring points p1 and p2, the temperature distribution difference value is the difference between the two temperature values. When the number of measuring points is n, (n-1) temperature distribution difference values are formed. According to these difference values, the temperature gradient change of each preset measuring point is determined by a spatial interpolation method. The temperature gradient change is represented as the temperature change rate per unit distance. The calculation of the temperature gradient change considers the two-dimensional spatial distribution of the measuring points, and the temperature gradients in the lateral and longitudinal directions are calculated respectively.

[0072] A mapping relationship between the temperature gradient change and the capacitance variation signal is established, which is realized by a lookup table. The temperature gradient change range of the lookup table is -5℃ / cm to 5℃ / cm with a step of 0.1℃ / cm, and the corresponding capacitance variation signal compensation coefficient range is 0.8 to 1.2. When the temperature gradient change exceeds the lookup table range, the boundary value is used for limitation. According to the mapping relationship, the temperature compensation coefficient is generated. For each preset measuring point, the corresponding compensation coefficient is found according to its temperature gradient change.

[0073] The generated temperature compensation coefficient is constructed into a temperature compensation correction matrix according to the positional relationship of the preset measurement points. Assuming that the preset measurement points on the cross arm surface form an m-row n-column matrix, the dimension of the temperature compensation correction matrix is also m-row n-column, and each element in the matrix corresponds to the temperature compensation coefficient of a preset measurement point. The temperature compensation correction matrix considers the spatial correlation between the measurement points, and performs smoothing processing on the compensation coefficients of adjacent measurement points with a smoothing radius of 2 measurement point spacings.

[0074] The cross arm surface temperature distribution has a spatial variation trend, and the temperature variation trends of the preset measurement points in the horizontal and vertical directions need to be calculated respectively. The horizontal temperature variation trend is obtained by linear regression of the temperature data of the same row of measurement points, and the regression coefficient represents the horizontal temperature variation rate. Similarly, the vertical temperature variation trend is obtained by linear regression of the temperature data of the same column of measurement points, and the regression coefficient represents the vertical temperature variation rate. The linear regression adopts the least squares method, and the regression coefficient is calculated again after removing outliers to improve the regression accuracy.

[0075] The horizontal and vertical temperature variation trends are combined with the temperature compensation correction matrix to construct a bidirectional compensation factor. The bidirectional compensation factor considers the comprehensive influence of temperature in the horizontal and vertical directions, and for a preset measurement point with a position of (i, j), its bidirectional compensation factor is the weighted average of the horizontal compensation factor and the vertical compensation factor, and the weight is dynamically adjusted according to the size of the temperature variation rate. When the horizontal temperature variation rate is greater than the vertical temperature variation rate, the weight of the horizontal compensation factor increases; otherwise, the weight of the vertical compensation factor increases. The weight range is 0 to 1, and the sum of the two is 1.

[0076] According to the bidirectional compensation factor, the capacitance change signal in the original capacitance data sequence is compensated in layers. The layered compensation considers the nonlinear influence of temperature on the capacitance signal, and is divided into two levels of coarse compensation and fine compensation. In the coarse compensation stage, the bidirectional compensation factor is used to directly correct the capacitance change signal to eliminate the systematic deviation caused by uneven temperature distribution. In the fine compensation stage, the compensation result is fine-tuned according to the local temperature fluctuation to adapt to the dynamic change of temperature. The compensation process adopts an iterative method, and the compensation effect is evaluated after each iteration. When the difference between the results of adjacent two iterations is less than a preset threshold, the iteration is ended. The preset threshold is set to 0.5% of the amplitude of the capacitance change signal.

[0077] After the temperature compensation processing, the capacitance data sequence obtained is compared with the original capacitance data sequence, and the interference caused by temperature change is eliminated, improving the accuracy and reliability of the signal. In the compensated capacitance data sequence, the capacitance change signal can more accurately reflect the actual state of the cross arm. The compensation result is evaluated by offline analysis and online verification. The offline analysis compares the capacitance change signals generated by the same displacement under different temperature conditions, and the online verification verifies the compensation effect by introducing a known displacement.

[0078] The application effectively solves the problem of interference of environmental temperature change on the capacitance signal by collecting the capacitance change signal of the cross arm through the capacitive sensor and performing temperature compensation processing. Based on the temperature distribution difference value, the temperature gradient change quantity is calculated, the compensation coefficient is generated combined with the mapping relationship of the capacitance change signal, the temperature compensation correction matrix is constructed, and then the transverse and longitudinal temperature change trends are analyzed to form the bidirectional compensation factor, thereby realizing the layered compensation of the capacitance signal. By comprehensively considering the spatial distribution characteristics and dynamic change law of temperature, this method realizes more fine and comprehensive temperature compensation effect, and provides more reliable technical support for cross arm state monitoring.

[0079] A multi-measurement-point mutual checking mechanism is established based on the capacitance data sequence after temperature compensation, multiple measurement points on the cross arm are selected, a constraint relationship between the capacitance data of each measurement point is constructed according to the mechanical equilibrium condition of the cross arm, and the theoretical capacitance value of each measurement point is calculated, including:

[0080] The trend analysis is performed on the capacitance data sequence after temperature compensation, the capacitance value change trend between adjacent measurement points on the cross arm is calculated, the capacitance change amplitude and change direction between the measurement points are extracted, the capacitance change correlation degree value between the measurement points is generated, and the capacitance change correlation matrix is constructed;

[0081] The capacitance change law between the measurement points is analyzed based on the capacitance change correlation matrix, the mutual influence degree between the measurement points is extracted, the measurement point capacitance data checking criterion is established, and the multi-measurement-point mutual checking mechanism is formed;

[0082] According to the multi-measurement-point mutual checking mechanism, multiple measurement points on the cross arm are selected to meet the stress transmission requirements, and the distribution positions of the measurement points are determined;

[0083] The stress transmission relationship between the multiple measurement points is calculated according to the mechanical equilibrium condition of the cross arm and the distribution positions of the measurement points, and the constraint relationship between the capacitance data of the multiple measurement points is constructed;

[0084] The distribution positions of the multiple measurement points are substituted into the constraint relationship between the capacitance data of the multiple measurement points to construct a constraint equation set, the constraint equation set is solved, and the theoretical capacitance value of each measurement point is calculated.

[0085] When performing trend analysis on the capacitance data sequence after temperature compensation, the sliding window method is used for processing, the window length is set to 30 sampling points, and the overlap rate is 50%. In each window, the capacitance value difference between adjacent measurement points on the cross arm is calculated to obtain the capacitance change trend. The change trend includes two characteristic parameters of change amplitude and change direction. The change amplitude is represented as the difference between the capacitance values of adjacent measurement points, and the change direction is represented as the positive or negative sign of the difference.

[0086] The extraction of the capacitance change trend adopts a double-threshold discrimination method, and the high threshold is set to 0.5 pF and the low threshold is set to 0.1 pF. When the capacitance change amplitude exceeds the high threshold, it is determined to be a significant change; when the capacitance change amplitude is between the high and low thresholds, it is determined to be a slight change; and when the capacitance change amplitude is lower than the low threshold, it is determined to be no obvious change. The change direction is determined according to the positive and negative of the difference value, and a positive value indicates an increase in the capacitance value, and a negative value indicates a decrease in the capacitance value. Based on the extracted change amplitude and change direction, the capacitance change correlation degree value between the measuring points is calculated. The calculation of the correlation degree value considers the proportional relationship of the change amplitude and the consistency of the change direction, and the correlation degree value ranges from 0 to 1, where 0 represents complete irrelevance and 1 represents complete relevance.

[0087] When constructing the capacitance change correlation matrix, the rows and columns of the matrix correspond to the measuring points on the cross arm respectively, and the matrix element value is the capacitance change correlation degree value between the corresponding two measuring points. The matrix is a symmetric matrix, and the main diagonal element value is 1, representing complete relevance of the measuring point to itself. In order to improve the stability of the matrix, the time average method is used to smooth the correlation degree value, and the time window is set to 10 minutes, and the average value of the correlation degree value in the window is taken as the final result.

[0088] Based on the constructed capacitance change correlation matrix, the capacitance change law between the measuring points is analyzed. The clustering analysis method is used to divide the measuring points with similar correlation degrees into the same category. The hierarchical clustering algorithm with a distance threshold of 0.2 is used for clustering, and the measuring point groups with high correlation degrees are gradually merged. For the clustering results, the mutual influence degree between the measuring points is extracted, and the influence degree is defined as a comprehensive index of the correlation degree and the distance between the measuring points. The closer the distance between two measuring points, the higher the correlation degree, and the greater the mutual influence degree. The influence degree is divided into high, medium and low levels, and the corresponding threshold values are 0.7 and 0.4.

[0089] Based on the mutual influence degree between the measuring points, the measuring point capacitance data verification criteria are established, and the verification criteria include the reasonable change range of the measuring point capacitance value and the consistency requirement of the capacitance change between the measuring points. For the measuring point pair with high influence degree, the capacitance change should have high consistency, and the allowed deviation range is 10%; for the measuring point pair with medium influence degree, the allowed deviation range is 20%; and for the measuring point pair with low influence degree, the allowed deviation range is 30%. The verification criteria are stored in matrix form, which is convenient for subsequent quick query and application.

[0090] The multi-point mutual verification mechanism is constructed based on verification criteria, including a measurement point selection strategy, an abnormality determination rule, and a data correction method. The measurement point selection strategy determines a set of measurement points participating in mutual verification, the abnormality determination rule is used to identify measurement points with abnormal capacitance data, and the data correction method is used to correct the capacitance values of abnormal measurement points. The mutual verification mechanism uses a majority voting method. When the capacitance value of a measurement point deviates from the expected values of the majority of related measurement points by more than the range specified in the verification criteria, it is determined that the data of the measurement point is abnormal.

[0091] According to the multi-point mutual verification mechanism, a plurality of measurement points that meet the stress transmission requirements are selected on the cross arm. The selection of the measurement points follows two principles: covering the key stress positions of the cross arm and ensuring the mutual verification capability between the measurement points. The key stress positions of the cross arm include the center point, the quarter point, and the support point. The stress changes at these positions can reflect the overall stress state of the cross arm. The number of measurement points is set to 9, which are distributed at typical positions on the surface of the cross arm. The coordinates of each measurement point are determined with the center of the cross arm as the origin, and the unit of the coordinates is centimeters. The measurement points are distributed in a grid shape, covering the main area of the cross arm, and ensuring that the overall stress distribution of the cross arm can be captured.

[0092] The cross arm is a stressed component, and its internal stress distribution follows the principles of elasticity. According to the geometric shape and support mode of the cross arm, a mechanical model is established. The stress types of the cross arm are mainly bending and torsion, and the corresponding stress distribution can be converted into the spatial distribution law of capacitance change. The mechanical equilibrium of the cross arm requires the balance of the force and moment of each part of the stress, which is converted into the constraint relationship between the capacitance values of each measurement point. The constraint relationship considers factors such as the position of the measurement point, the stiffness of the cross arm, and the stress transmission efficiency, forming a linear combination equation of the capacitance values of the measurement points.

[0093] The constraint relationship forms a constraint equation set, and the number of equations is comparable to the number of measurement points, ensuring that the equation set has a definite solution. The solution of the constraint equation set uses an iterative method, and the initial value is set to the actual measured capacitance value. In the iteration process, the capacitance values of each measurement point are continuously adjusted according to the constraint conditions until all constraint conditions are satisfied, or the iteration number reaches the preset upper limit of 100 times, or the difference between the results of two consecutive iterations is less than 0.01 pF. The capacitance values of each measurement point obtained by solving are the theoretical capacitance values, which are used as the reference values for subsequent data anomaly detection and correction.

[0094] When the measured capacitance value of a measurement point deviates from the theoretical capacitance value by more than a preset threshold, it is determined that the data of the measurement point is abnormal. The threshold is set to 15% of the theoretical capacitance value, which can be adjusted according to actual needs. For the measurement points determined to be abnormal, the theoretical capacitance value is used to replace the abnormal value, or the theoretical capacitance value and the historical data are used for weighted correction with weight coefficients of 0.7 and 0.3. The corrected data is re-verified in the multi-point mutual verification mechanism to ensure that the correction results meet the mechanical equilibrium conditions of the cross arm.

[0095] The application significantly improves the reliability and accuracy of the capacitance data by establishing mutual checking mechanism between multiple measuring points on the cross arm. The mutual checking mechanism of multiple measuring points fully utilizes the mechanical characteristics of the cross arm and the spatial distribution characteristics of the measuring points, realizes the self-calibration and self-correction functions of the capacitance data, and enhances the robustness and adaptability. Through the comparison and analysis of the theoretical capacitance value and the measured value, not only the data anomaly can be monitored, but also the stress state change of the cross arm structure can be reflected, which provides more reliable data support for the state evaluation and early warning of the cross arm.

[0096] The deviation of the actual capacitance value from the theoretical capacitance value is taken as the correction amount to compensate the data of each measuring point, and the mutual checking capacitance data sequence is obtained.

[0097] The actual capacitance value of each measuring point is obtained, the difference sequence between the actual capacitance value and the theoretical capacitance value is calculated, and the measuring point capacitance deviation sequence is generated according to the difference sequence.

[0098] The stress transmission path is determined according to the stress characteristics of the cross arm, the change characteristics of the measuring point capacitance deviation sequence are analyzed along the stress transmission path, the deviation change amount and the transmission law between adjacent measuring points are extracted, and the measuring point compensation parameter is constructed.

[0099] The spatial distribution of the deviation change amount is corrected by using the measuring point compensation parameter, and the dynamic compensation coefficient is generated based on the transmission law.

[0100] The dynamic compensation coefficient is used to modify the actual capacitance measurement value of each measuring point in sequence along the stress transmission path, and the mutual checking capacitance data sequence is obtained.

[0101] The actual capacitance value of each measuring point is obtained, which is obtained by real-time collection of the capacitance sensor arranged on the cross arm. The sampling frequency of the capacitance sensor is set to 20Hz, and the sampling accuracy is 0.01pF. The collected capacitance data is processed by temperature compensation and baseline calibration to form the actual capacitance value sequence. The theoretical capacitance value is calculated by the constraint equation set constructed by the mechanical balance condition of the cross arm, which represents the capacitance value of each measuring point under normal stress state. The actual capacitance value and the theoretical capacitance value are compared and calculated to obtain the difference sequence between them, which records the capacitance deviation of each measuring point at different time points.

[0102] The calculation of the difference sequence adopts the point-by-point comparison method, and the actual capacitance value at each time point is subtracted from the corresponding theoretical value to form time series data. The difference sequence is smoothed by a sliding window, and the window length is set to 15 sampling points. The smoothed data is used as the capacitance deviation sequence of the measuring point. The capacitance deviation sequence reflects the deviation degree of the capacitance value of each measuring point relative to the theoretical value. The positive and negative signs of the deviation sequence indicate that the capacitance value is high or low. The absolute value of the deviation sequence indicates the degree of deviation. The deviation sequence is also filtered by a threshold to remove small fluctuations. The threshold is set to 0.05 pF, and the deviation below the threshold is considered as normal fluctuation and ignored.

[0103] As a key stressed component in the power transmission line, the stress distribution of the cross arm has a clear transmission rule. According to the material properties and geometric shape of the cross arm, combined with the finite element analysis results, the main path of stress transmission on the cross arm is determined. The stress transmission path usually starts from the stress point and spreads to both ends along the cross arm structure, forming a tree-like or network-like transmission network. The stress relationship between adjacent nodes on the stress transmission path shows obvious correlation, which can be used to guide the correction of the capacitance deviation. The stress transmission path is divided into a main path and a secondary path according to the influence intensity. The main path covers the most significant part of the cross arm stress, and the secondary path covers the less significant part of the stress.

[0104] When analyzing the variation characteristics of the capacitance deviation sequence of the measuring point along the stress transmission path, the amplitude variation, propagation delay and attenuation law of the deviation sequence are mainly concerned. The amplitude variation reflects the difference in the deviation between adjacent measuring points, the propagation delay reflects the time required for the stress wave to transmit from one measuring point to another, and the attenuation law reflects the weakening degree of the stress in the transmission process. By comparing and analyzing the deviation sequences of adjacent measuring points, the deviation variation is extracted, which is represented as the difference between the capacitance deviations of adjacent measuring points. The deviation variation, combined with the spatial distance and time delay between the measuring points, forms the deviation transmission characteristics between the measuring points.

[0105] The deviation transmission law between the measuring points includes a spatial attenuation coefficient and a time delay parameter. The spatial attenuation coefficient describes the weakening degree of the stress when it is transmitted from one measuring point to an adjacent measuring point, which is usually related to the distance between the measuring points and the material properties of the cross arm. The time delay parameter describes the time required for the stress wave to propagate between the measuring points, which is usually related to the distance between the measuring points and the elastic wave speed of the cross arm material. According to the extracted deviation variation and the transmission law, the measuring point compensation parameter is constructed. The measuring point compensation parameter includes a spatial correction factor and a time calibration factor, which are used to correct the capacitance deviation of the measuring point.

[0106] When using the compensation parameters of the measuring points to correct the spatial distribution of the deviation variation, the spatial interpolation method is used to fill in the deviation variation between the measuring points. For any position between two adjacent measuring points, the deviation variation can be calculated through the spatial correction factor and the deviation variation of the two end measuring points. The spatial distribution correction considers the spatial non-uniformity of the measuring points, and different interpolation strategies are used for sparse areas and dense areas. Linear interpolation is used for sparse areas, and weighted average interpolation is used for dense areas, and the weight is inversely proportional to the distance between the measuring points. After spatial distribution correction, the deviation variation forms a continuous spatial field, which more accurately reflects the stress distribution on the cross arm.

[0107] The dynamic compensation coefficient adjusts with the change of the stress on the cross arm, and adapts to the deviation correction requirements under different working conditions. The calculation of the dynamic compensation coefficient is based on the time calibration factor and the historical data of the measuring points, and the exponential weighted average method is used to integrate information of multiple time scales. The dynamic compensation coefficient is divided into steady-state coefficient and transient coefficient, the steady-state coefficient is used for long-term slow change deviation correction, and the transient coefficient is used for sudden change deviation correction. The update frequency of the dynamic compensation coefficient is consistent with the capacitance sampling frequency, ensuring that the compensation follows the change of the stress on the cross arm in time.

[0108] When the dynamic compensation coefficient is used to correct the actual capacitance measurement value of each measuring point in sequence according to the stress transmission path, the correction starting point and the correction direction need to be determined. The correction starting point selects the most reliable measuring point on the cross arm, usually the measuring point with the smallest deviation verified by multiple times. Starting from the correction starting point, the measuring points are corrected in topological order along the stress transmission path. The correction calculation of each measuring point considers the correction results of the upstream measuring points and the dynamic compensation coefficient of the measuring point, and the correction formula is the actual capacitance value minus the corrected deviation. The correction calculation uses progressive processing, and the correction result of the previous measuring point affects the correction calculation of the subsequent measuring points, forming a cascading effect.

[0109] When there are multiple stress transmission paths on the cross arm, the correction calculation needs to consider the special processing of the path intersection point. The path intersection point receives the influence of multiple upstream measuring points, and its correction amount needs to be integrated with the correction results of multiple paths. The multi-path integration uses the weighted average method, and the weight is proportional to the path reliability and influence strength. The path reliability is verified and evaluated through historical data, and the influence strength is determined through stress analysis. For possible correction conflicts, a conflict resolution mechanism is set to preferentially use the correction result of the path with high reliability.

[0110] The mutual checking after the above-mentioned correction calculation obtains the capacitance data sequence, as the basis data for subsequent cross arm state evaluation. In order to ensure the reliability of the correction result, a correction effectiveness verification mechanism is introduced. The correction effectiveness is evaluated by comparing the consistency and physical rationality of the data before and after the correction, the consistency evaluation adopts the correlation coefficient and mean square deviation index, and the physical rationality evaluation is based on the stress principle and structural characteristics of the cross arm. The stress principle considers the stress distribution characteristics of the cross arm under various working conditions, including the vertical force generated by the suspended conductor, the horizontal tension and the complex stress state caused by the comprehensive factors such as wind load. When the correction effectiveness verification fails, an alarm is triggered and the original data is returned to avoid the judgment error caused by the wrong correction.

[0111] The application extracts the deviation transmission law between adjacent measuring points by analyzing the change characteristics of the measuring point capacitance deviation sequence, and establishes a compensation mechanism suitable for the dynamic stress change of the cross arm. The strategy of point-by-point correction along the stress transmission path fully utilizes the mechanical characteristics of the cross arm structure, so that the correction process has physical meaning. The introduction of the dynamic compensation coefficient enhances the adaptability of the correction, which can cope with the deviation change under different working conditions. The mutual checking after the capacitance data sequence has higher precision and stability, which provides a reliable data basis for the cross arm state monitoring and fault diagnosis, and effectively improves the safety and reliability of the power transmission line operation.

[0112] The mutual checking after the capacitance data sequence is segmented and converted into stress values to obtain a preliminary stress data sequence, and the preliminary stress data sequence is converted back to capacitance data to obtain a reconstructed capacitance data sequence, including:

[0113] The capacitance change trend of the measuring point in the mutual checking after the capacitance data sequence is analyzed, the capacitance change amplitude between adjacent measuring points is calculated, the mutation position is identified according to the capacitance change amplitude, and the data segmentation interval is determined;

[0114] The stress deformation characteristics of the measuring point in the data segmentation interval are extracted, the corresponding relationship between the stress deformation characteristics and the capacitance change is analyzed, the capacitance-stress conversion coefficient of each data segmentation interval is generated, and the segmented mapping relationship from capacitance to stress is constructed;

[0115] The segmented mapping relationship is used for conversion calculation on the mutual checking after the capacitance data sequence, and the numerical smooth connection at the segmented boundary is obtained to obtain a preliminary stress data sequence;

[0116] The capacitance-stress conversion coefficient is reversely configured to establish a reverse calculation relationship from stress to capacitance, and the preliminary stress data sequence is reversely converted to obtain a reconstructed capacitance data sequence.

[0117] The capacitance data sequence after mutual verification is high-reliability capacitance measurement data obtained through pre-processing, which contains the capacitance change information of each measuring point of the cross arm under different working conditions. Analyzing the change trend of the capacitance data sequence can identify the law of capacitance change with stress. The capacitance change trend analysis adopts a differential method to process the capacitance value difference between adjacent time points, forming a capacitance change rate sequence. The capacitance change amplitude between adjacent measuring points is obtained by comparing the capacitance change rate sequences of different measuring points, and is expressed as the difference value of the capacitance change rate between measuring points.

[0118] The capacitance change amplitude sequence reflects the spatial variation characteristics of the stress distribution on the cross arm. By setting a capacitance change amplitude threshold, positions with a change amplitude exceeding the threshold can be identified, which are defined as capacitance mutation positions. The mutation positions usually correspond to regions with significant changes in the stress state of the cross arm, such as stress concentration areas or load transfer areas. The capacitance change amplitude threshold is set according to historical data statistics and engineering experience, and is generally taken as twice the average value of the capacitance change amplitude, which can be adjusted according to actual application scenarios. According to the identified mutation positions, the capacitance data sequence is divided into several segmented intervals, and the capacitance change characteristics in each segmented interval are relatively consistent.

[0119] After determining the data segmented intervals, the stress deformation characteristics of the measuring points are extracted in each interval. The stress deformation characteristics include key information such as linear intervals, saturation intervals, and transition intervals, reflecting the mechanical response characteristics of the cross arm under different load levels. By analyzing the distribution characteristics of the capacitance data in each segmented interval, the concentration trend and dispersion degree of the data are determined, and abnormal data points are identified and filtered. The abnormal data judgment standard is data points deviating from the interval mean value by more than 3 times the standard deviation. The filtered data can more accurately reflect the stress deformation characteristics of the cross arm.

[0120] The corresponding relationship between the stress deformation characteristics and the capacitance change is established through data matching analysis. For each segmented interval, the capacitance measurement values under known load conditions are collected, and the stress measurement values at the corresponding positions are recorded to form capacitance-stress data pairs. The capacitance-stress data pairs are processed through regression analysis to establish the mapping relationship between the capacitance value and the stress value. Regression analysis adopts a segmented fitting strategy, using different fitting curves for different load ranges to adapt to the nonlinear change characteristics of the capacitance-stress relationship under different load levels.

[0121] According to the regression analysis results, the capacitance-stress conversion coefficients of each data segment interval are generated. The capacitance-stress conversion coefficient includes a proportional coefficient and an offset, which correspond to the slope and intercept of the capacitance-stress mapping relationship, respectively. For a linear interval, the proportional coefficient remains a constant value; for a nonlinear interval, the proportional coefficient is dynamically adjusted with the change of the capacitance value. The accuracy of the capacitance-stress conversion coefficient directly affects the accuracy of stress calculation. In the coefficient generation process, multiple calibration and verification procedures are used to ensure the reliability of the conversion coefficient.

[0122] After the construction of the segmented mapping relationship of capacitance to stress, the mapping relationship is used to convert and calculate the capacitance data sequence after mutual verification. In the conversion calculation process, for the capacitance value of each measurement point, the corresponding stress value is calculated according to the segmented interval it is located in and the selection of the corresponding capacitance-stress conversion coefficient. When the capacitance value is near the segmented boundary, a smooth transition strategy is used to handle the segmented boundary to avoid sudden changes in the calculation results. The smooth transition uses a weighted average method, and the calculation results on both sides of the boundary are weighted and averaged according to the distance from the boundary, and the weight is proportional to the distance from the boundary.

[0123] The numerical smooth connection at the segmented boundary ensures the continuity of the stress data. For the junction of adjacent segments, there is a risk of discontinuity in the calculation results, which may cause unnatural jumps in the stress distribution. To solve this problem, a transition area is set at the boundary, and the width of the transition area is generally 5% to 10% of the length of the segmented interval. In the transition area, the stress value is smoothly transitioned through linear interpolation to ensure the continuity and smoothness of the stress data in the entire interval. The stress data obtained through the above processing is referred to as the preliminary stress data sequence, which reflects the stress distribution state of the cross arm under the current load condition.

[0124] To verify the accuracy of the preliminary stress data sequence, the capacitance-stress conversion coefficient needs to be configured in reverse to establish the reverse calculation relationship of stress to capacitance. In the reverse configuration process, the independent variable and the dependent variable in the capacitance-stress mapping relationship are exchanged to form the stress-capacitance mapping relationship. The establishment of the reverse calculation relationship also uses a segmented strategy, and different conversion parameters are set for different stress level intervals. The accuracy of the reverse calculation relationship matches the forward conversion coefficient to ensure the reversibility of the conversion.

[0125] The preliminary stress data sequence is reversely converted by using the reverse calculation relationship of stress to capacitance to obtain a reconstructed capacitance data sequence. The calculation process of reverse conversion is similar to forward conversion, and the corresponding capacitance value is calculated according to the interval of the stress value to select the corresponding conversion parameter. The deviation between the reverse conversion result and the original mutual check capacitance data sequence is calculated, and the deviation size reflects the accuracy loss of the conversion process. The deviation threshold is set to 1% of the original capacitance value, and the measurement points exceeding the threshold need to be recalculated by adjusting the conversion parameter.

[0126] The consistency verification of the reconstructed capacitance data sequence and the original mutual check capacitance data sequence is the effectiveness of the whole conversion process. The consistency verification is evaluated by multiple indicators, including correlation coefficient, root mean square error and the degree of agreement of capacitance change trend. The reconstructed capacitance data sequence that passes the verification can be used as a reliable basis for the evaluation of the stress state of the cross arm, and provides data support for subsequent structural safety analysis.

[0127] The application establishes a two-way mapping mechanism between the capacitance measurement value and the cross arm stress state by converting the mutual check capacitance data sequence into stress values in segments and performing reverse verification, realizes accurate conversion from the capacitance domain to the stress domain. The segmented processing strategy effectively solves the nonlinear change problem of the capacitance-stress relationship under different load levels, improves the adaptability and accuracy of the conversion. The smooth connection at the segment boundary ensures the continuity of the stress distribution, avoids the misjudgment caused by unnatural jump. The reverse conversion mechanism provides a self-verification means for the conversion process, which significantly enhances the reliability of the stress calculation result.

[0128] As shown in Figure 2 , the residual error of the reconstructed capacitance data sequence and the mutual check capacitance data sequence is calculated, the spatial gradient of the residual error along the measurement point position is extracted, the mutation measurement point is identified, the conversion relationship of the adjacent measurement points is used to correct the mutation measurement point, and an optimized cross arm stress data sequence is obtained, including:

[0129] The numerical difference of the reconstructed capacitance data sequence and the mutual check capacitance data sequence at the measurement point position is calculated, the measurement point residual data is generated, and the residual change amount between the measurement points is calculated according to the measurement point residual data;

[0130] The residual distribution function is constructed by using the measurement point residual data and the residual change amount, the spatial direction derivative of the residual distribution function at the measurement point position is calculated, and the residual spatial gradient value is obtained;

[0131] The measurement point position where the residual spatial gradient value exceeds the preset gradient range is determined as the mutation measurement point, the capacitance change amount of the measurement points on both sides of the mutation measurement point is extracted, and the measurement point conversion function is generated;

[0132] The point conversion function is applied to the stress data at the mutation point to correct and compensate, so as to obtain an optimized cross arm stress data sequence.

[0133] When calculating the residual of the reconstructed capacitance data sequence and the mutual checking capacitance data sequence, the two sequences need to be compared in detail. The reconstructed capacitance data sequence is the capacitance estimated value obtained by reverse conversion of the stress data, and the mutual checking capacitance data sequence is the measured capacitance value after the previous processing. In an ideal case, the two sequences should be highly consistent, but in actual operation, differences may occur due to factors such as conversion model accuracy limitations, noise interference, etc. The measurement point residual data is the difference between the reconstructed capacitance value and the measured capacitance value. The measurement point residual data reflects the error degree in the capacitance-stress conversion process, and the smaller the absolute value, the more accurate the conversion, and the larger the error, the more attention needs to be paid to the conversion relationship at this point.

[0134] The residual variation between measurement points refers to the variation amplitude of residual data between adjacent measurement points. The residual variation can reflect the spatial variation characteristics of the residual distribution, and the position with a larger variation usually corresponds to the mutation area of the residual distribution. For uniformly distributed measurement points on the cross arm, the residual variation between adjacent measurement points should be relatively stable in theory. When the residual variation between measurement points exceeds three times the standard deviation of the statistical average, it can be preliminarily determined that there is an anomaly in this area. The calculation of the residual variation needs to consider the spatial distribution characteristics of the measurement points. For measurement points with uneven spacing, the residual variation can be converted to a variation rate per unit distance through normalization processing.

[0135] When constructing the residual distribution function using the measurement point residual data and the residual variation, the interpolation method is used to expand the discrete measurement point residual data into a continuous spatial distribution function. Common interpolation methods include linear interpolation, spline interpolation, etc., among which cubic spline interpolation can ensure the smoothness of the curve. In the interpolation process, the spatial position of the measurement point is taken as the independent variable, and the measurement point residual data is taken as the dependent variable, to generate a residual distribution function covering the entire cross arm area. The construction of the residual distribution function needs to consider the setting of the boundary condition, and the natural boundary condition is generally used to ensure smooth transition of the residual distribution.

[0136] After the residual distribution function is constructed, the spatial directional derivative of the function at the measurement point position is calculated to obtain the residual spatial gradient value. The spatial directional derivative represents the variation rate of the residual in space and can be calculated by difference approximation. Central difference has high accuracy and is suitable for gradient calculation at most measurement point positions; while for boundary measurement points, forward difference or backward difference can be used for processing. The unit of the residual spatial gradient value is related to the measurement point spacing, and to ensure the comparability between different regions, the gradient value is usually normalized.

[0137] When analyzing the residual space gradient value to identify the abrupt measurement point, a preset gradient range needs to be set as the determination standard. The preset gradient range is usually determined based on historical data statistics and professional knowledge, and it is desirable to have a statistical average value of the residual space gradient value under normal working conditions floating up and down by a certain percentage. When the residual space gradient value of a certain measurement point exceeds the preset gradient range, the measurement point is marked as an abrupt measurement point. The determination of the abrupt measurement point also needs to combine the gradient distribution characteristics of the surrounding area of the measurement point to avoid misjudgment caused by local noise. Generally, the gradient value of the abrupt measurement point is required to exceed the preset range for at least three consecutive sampling points to be confirmed as an effective mutation.

[0138] After the abrupt measurement point is determined, the capacitance change amount of the measurement points on both sides of the measurement point is extracted, and the change rule is analyzed. The capacitance change amount refers to the change amplitude of the measurement point capacitance value relative to the reference state, which reflects the stress change of the cross arm at this point. Usually, three to five measurement points on both sides of the abrupt measurement point are selected as the reference interval, and the capacitance change amount sequence of these measurement points under different load levels is extracted. By analyzing the distribution characteristics of the capacitance change amount in the reference interval, the cause of the abnormality of the abrupt measurement point can be judged.

[0139] The measurement point conversion function is constructed based on the capacitance-stress corresponding relationship of the normal measurement points on both sides of the abrupt measurement point, and a weighted interpolation method is used to integrate the conversion characteristics of the measurement points on both sides. The weight coefficient is inversely proportional to the distance of the measurement point to the abrupt position, and the closer the measurement point, the greater the contribution to the conversion function. For different types of mutation reasons, the construction strategy of the measurement point conversion function is different.

[0140] The measurement point conversion function is applied to the stress data at the abrupt measurement point for correction and compensation to obtain an optimized cross arm stress data sequence. The correction process is based on the original stress data, and the correction value is calculated according to the measurement point conversion function. To ensure smooth transition, a transition area is set near the abrupt measurement point, and the correction amount in the transition area gradually decreases as the distance to the abrupt measurement point increases, until it seamlessly connects with the original stress data. The stress data after correction and compensation needs to be verified for effectiveness, and the verification methods include re-computing the residual distribution to check whether it is smooth, comparing with historical data to analyze the consistency of the trend, etc.

[0141] The present application accurately identifies data abnormal points through residual space gradient analysis, and uses the conversion relationship of adjacent measurement points for targeted correction. It avoids the problem of possible global error accumulation, maintains the accurate expression of local characteristics, and effectively improves the accuracy and reliability of cross arm stress monitoring. The optimization process does not require additional sensor deployment, fully utilizes existing measurement data for self-correction, and reduces the implementation cost. The residual distribution function construction, spatial gradient calculation and mutation measurement point correction strategy in the method form a complete data optimization closed loop, so that the cross arm stress monitoring can still maintain stable and reliable performance under complex working conditions, providing more accurate technical support for the safe operation of power transmission lines.

[0142] Based on the optimized cross arm stress data, the bearing state of the cross arm is evaluated, the stress abnormal area is identified, and the early warning information is generated, which includes:

[0143] The stress distribution characteristics of the measuring points in the optimized cross arm stress data sequence are analyzed, the stress change trend of the measuring points is extracted, and the cross arm stress distribution atlas is constructed;

[0144] According to the cross arm stress distribution atlas, the stress concentration degree of each region of the cross arm is calculated, the stress state of the cross arm is evaluated, and the cross arm bearing state evaluation result is generated;

[0145] Based on the cross arm bearing state evaluation result, the stress super limit region is identified, the stress aggregation degree in the stress super limit region is calculated, and the stress risk assessment index is constructed;

[0146] The stress risk assessment index is combined with the stress aggregation degree for hierarchical judgment, and the early warning information containing the risk level and the region positioning is generated.

[0147] When analyzing the stress distribution characteristics of the measuring points in the optimized cross arm stress data sequence, spatial correlation analysis of the stress values of each measuring point is required. The cross arm measuring points are usually arranged along the length direction of the cross arm, and the measuring point spacing is generally 1 / 10 to 1 / 20 of the length of the cross arm, to ensure sufficient spatial resolution. The stress distribution characteristics include maximum stress value, minimum stress value, average stress level, stress gradient and other parameters. The stress change trend of the measuring points needs to be extracted by time domain analysis of the stress data sequence, and the sliding window method can be used to calculate the change rate of the stress values of each measuring point. The sliding window size is recommended to be set to 5 minutes to 30 minutes, and the window overlap rate is 50% to 80%, to balance the calculation efficiency and trend sensitivity. The stress change trend can be divided into three basic types: rising, falling and stable, and when the stress change rate in the continuous three windows exceeds the preset threshold (recommended value is 2 times of the standard deviation), it is determined as an effective trend change.

[0148] Constructing the cross arm stress distribution atlas is the process of converting the stress data of discrete measuring points into continuous spatial distribution expression. The stress distribution atlas is represented in two dimensions, with the horizontal axis as the cross arm spatial position coordinate and the vertical axis as the stress value. For the region between measuring points, an interpolation algorithm is used to generate a continuous stress distribution curve, and common interpolation methods include linear interpolation, spline interpolation, etc. High-order spline interpolation is suitable for sparse measuring points, and can generate smooth stress distribution curves; while piecewise linear interpolation is suitable for dense measuring point regions, and has higher calculation efficiency. The atlas update frequency should match the data acquisition frequency, and is generally updated once every 5 to 10 minutes.

[0149] When calculating the stress concentration degree of each region of the cross arm according to the stress distribution map of the cross arm, the stress concentration degree calculation method needs to be defined. The stress concentration degree is a deviation degree of the stress level of a local region relative to the average stress level of the whole, which can be represented by the ratio of the average stress of the local region to the average stress of the whole. The cross arm can be divided into several sub-regions for stress concentration degree calculation. The principle of sub-region division is to ensure that each sub-region contains at least two measuring points and the length is not more than 1 / 5 of the total length of the cross arm. The division of sub-regions can be equidistant division or functional division based on the structural characteristics of the cross arm. When the stress concentration degree of a certain region exceeds 1.5, it indicates that there is stress concentration in that region; when the stress concentration degree exceeds 2.0, it is a significant stress concentration, which needs to be paid special attention to.

[0150] The evaluation of the stress state of the cross arm is a comprehensive judgment process based on the stress concentration degree, combined with historical data and theoretical models. The stress state of the cross arm can be divided into four levels: normal, mild abnormal, moderate abnormal and severe abnormal. The normal state means that the stress concentration degree of each region does not exceed 1.5, and the stress distribution is typically bimodal or unimodal; the mild abnormality means that there is a region with stress concentration degree between 1.5 and 2.0, but the overall shape of the stress distribution remains relatively stable; the moderate abnormality means that there is a region with stress concentration degree exceeding 2.0, or the shape of the stress distribution changes significantly; the severe abnormality means that there is a region with stress concentration degree exceeding 3.0, or multiple high stress regions are connected to each other. The evaluation results of the cross arm bearing state should include the stress state level, the position identification of the abnormal region, the stress overrun percentage and other key information.

[0151] When identifying the stress overrun region based on the evaluation results of the cross arm bearing state, the stress overrun judgment standard needs to be set. The stress overrun judgment standard should consider the strength characteristics of the cross arm material and the safety margin requirement, and generally take 70% to 80% of the yield strength of the material as the stress overrun threshold. For composite cross arms, the anisotropic properties of the material also need to be considered, and the stress overrun thresholds for longitudinal and transverse directions are set respectively. The identification of stress overrun region uses the threshold screening method, which marks the regions exceeding the threshold in the stress distribution map as stress overrun regions. The boundary of the stress overrun region is determined by the gradient tracking method, which expands from the overrun point to the surrounding until the stress gradient change rate is lower than the preset value (the recommended value is 20% of the average gradient).

[0152] The stress aggregation degree in the stress overrun area is calculated to quantitatively evaluate the severity of the overrun. The stress aggregation degree considers three factors of the spatial range, the overrun amplitude and the duration of the stress overrun area. The spatial range factor refers to the proportion of the stress overrun area in the total length of the cross arm, and the larger the range, the higher the aggregation degree. The overrun amplitude factor refers to the percentage of the average stress value in the stress overrun area exceeding the threshold value, and the more the overrun, the higher the aggregation degree. The duration factor refers to the length of time of the stress overrun, and the longer the time, the higher the aggregation degree. The three factors are comprehensively calculated by a weighted average method to obtain the final stress aggregation degree, and the weight proportion is suggested to be spatial range:overrun amplitude:duration=3:5:2. The stress aggregation degree has a value range of 0 to 100, and the larger the value, the higher the risk.

[0153] In constructing the stress risk evaluation index, the stress aggregation degree is combined with the historical bearing capacity of the cross arm and the environmental condition influencing factors. The historical bearing capacity of the cross arm reflects the fatigue degree and cumulative damage condition of the cross arm, which can be calculated by analyzing the historical stress cycle times and amplitude. The environmental condition influencing factors mainly include temperature, humidity, wind speed, etc., and extreme environmental conditions can reduce the actual bearing capacity of the cross arm. The calculation of the stress risk evaluation index adopts a multi-factor comprehensive scoring method, and the weight of each factor is determined according to actual engineering experience. The evaluation index has a value range of 0 to 100, and 60 is generally taken as the starting threshold of risk warning.

[0154] When the stress risk evaluation index is combined with the stress aggregation degree for grading judgment, a four-level risk grade division standard is established: normal, attention, warning and danger. The normal level corresponds to the risk evaluation index less than 60 and the stress aggregation degree less than 40; the attention level corresponds to the risk evaluation index between 60 and 75 or the stress aggregation degree between 40 and 60; the warning level corresponds to the risk evaluation index between 75 and 90 or the stress aggregation degree between 60 and 80; and the danger level corresponds to the risk evaluation index greater than 90 or the stress aggregation degree greater than 80. The generated warning information includes the risk grade, the positioning coordinates of the overrun area, the predicted development trend, etc. The warning information can be pushed to the operation and maintenance personnel through SMS, email, etc., and displayed in a conspicuous way on the monitoring interface.

[0155] The present application realizes the accurate evaluation and warning of the bearing state of the cross arm by analyzing the stress distribution characteristics of the cross arm in multiple dimensions. Based on the optimized stress data sequence of the cross arm, a complete stress analysis and processing flow is constructed from stress distribution map construction to stress concentration degree calculation, and then to risk grade judgment and warning information generation. The stress aggregation degree calculation method and risk grading standard adopted effectively balance the sensitivity and reliability, avoiding false positives and false negatives. The regional positioning function of the warning information provides accurate guidance for subsequent maintenance, saving maintenance resources.

[0156] The embodiment of the application provides a cross arm capacitive stress data precision optimization system, and the system comprises:

[0157] A data acquisition and temperature compensation unit is configured to acquire a capacitive change signal of the cross arm through a capacitive sensor, obtain an original capacitive data sequence, and perform temperature compensation processing on the original capacitive data sequence to obtain a temperature-compensated capacitive data sequence.

[0158] A multi-measurement-point mutual verification unit is configured to establish a multi-measurement-point mutual verification mechanism based on the temperature-compensated capacitive data sequence, select multiple measurement points on the cross arm, construct a constraint relationship between capacitive data of each measurement point according to a mechanical balance condition of the cross arm, and calculate a theoretical capacitive value of each measurement point.

[0159] A data compensation unit is configured to compensate measurement point data by taking a deviation between an actual capacitive value and the theoretical capacitive value as a correction amount, and obtain a mutual-verification capacitive data sequence.

[0160] A data conversion unit is configured to convert the mutual-verification capacitive data sequence into stress values in segments to obtain a preliminary stress data sequence, convert the preliminary stress data sequence back into capacitive data, and obtain a reconstructed capacitive data sequence.

[0161] A mutation measurement point correction unit is configured to calculate a residual error between the reconstructed capacitive data sequence and the mutual-verification capacitive data sequence, extract a spatial gradient of the residual error along a measurement point position, identify a mutation measurement point, correct the mutation measurement point by using a conversion relationship of adjacent measurement points, and obtain an optimized cross arm stress data sequence.

[0162] An early warning unit is configured to evaluate a bearing state of the cross arm based on the optimized cross arm stress data, identify a stress abnormal area, and generate early warning information.

[0163] The above-described specific embodiments are preferred embodiments of the application, and the specific implementation range of the application is not limited by the above-described specific embodiments. The scope of the application includes but is not limited to the specific embodiments, and equivalent changes made according to the shape and structure of the application are within the protection scope of the application.

Claims

1. A cross arm capacitive stress data accuracy optimization method, characterized in that, The method comprises the following steps: Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence; Based on the temperature-compensated capacitance data sequence, a multi-measurement-point mutual verification mechanism is established, a plurality of measurement points on the cross arm are selected, and a constraint relationship between the capacitance data of each measurement point is constructed according to the mechanical equilibrium condition of the cross arm to calculate the theoretical capacitance value of each measurement point; The deviation of the actual capacitance value from the theoretical capacitance value is taken as a correction amount to compensate the data of each measurement point, and a mutual-verified capacitance data sequence is obtained; The mutual-verified capacitance data sequence is segmented and converted into a stress value to obtain a preliminary stress data sequence, and the preliminary stress data sequence is converted back into capacitance data to obtain a reconstructed capacitance data sequence; The residual error between the reconstructed capacitance data sequence and the mutual-verified capacitance data sequence is calculated, the spatial gradient of the residual error along the measurement point position is extracted, the abrupt measurement point is identified, the conversion relationship of adjacent measurement points is used to correct the abrupt measurement point, and an optimized cross arm stress data sequence is obtained; Based on the optimized cross arm stress data, the bearing state of the cross arm is evaluated, the stress abnormal area is identified, and a warning information is generated.

2. The method of claim 1, wherein, Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence includes: Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence includes: Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence includes: Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence includes: Collecting the capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and performing temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence includes: Based on the temperature-compensated capacitance data sequence, a multi-measurement-point mutual verification mechanism is established, a plurality of measurement points on the cross arm are selected, and a constraint relationship between the capacitance data of each measurement point is constructed according to the mechanical equilibrium condition of the cross arm to calculate the theoretical capacitance value of each measurement point includes:

3. The method of claim 1, wherein, Performing trend analysis on the temperature-compensated capacitance data sequence, calculating the capacitance value change trend between adjacent measurement points on the cross arm, extracting the capacitance change amplitude and direction between the measurement points, generating the capacitance change correlation degree value between the measurement points, and constructing a capacitance change correlation matrix; Based on the capacitance change correlation matrix, the capacitance change law between the measurement points is analyzed, the mutual influence degree between the measurement points is extracted, a measurement point capacitance data verification criterion is established, and a multi-measurement-point mutual verification mechanism is formed; According to the multi-measurement-point mutual verification mechanism, a plurality of measurement points on the cross arm that meet the stress transmission requirements are selected, and the distribution positions of the measurement points are determined; ​ According to the mechanical equilibrium condition of the cross arm and the distribution positions of the measuring points, a stress transmission relationship between the measuring points is calculated, and a constraint relationship between the capacitance data of the measuring points is constructed; The distribution positions of the measuring points are substituted into the constraint relationship between the capacitance data of the measuring points to construct a constraint equation group, the constraint equation group is solved, and the theoretical capacitance values of the measuring points are calculated.

4. The method of claim 1, wherein, The deviation between the actual capacitance values and the theoretical capacitance values is taken as a correction amount to compensate the data of the measuring points, and a capacitance data sequence after mutual checking is obtained, including: Actual capacitance values of the measuring points are obtained, a difference sequence between the actual capacitance values and the theoretical capacitance values is calculated, and a measuring point capacitance deviation sequence is generated according to the difference sequence; According to the stress transmission path of the cross arm, the variation characteristics of the measuring point capacitance deviation sequence are analyzed along the stress transmission path, the deviation variation amount and the transmission law between adjacent measuring points are extracted, and a measuring point compensation parameter is constructed; The spatial distribution of the deviation variation amount is corrected by using the measuring point compensation parameter, and a dynamic compensation coefficient is generated based on the transmission law; The dynamic compensation coefficient is sequentially used to correct and calculate the actual capacitance measurement values of the measuring points along the stress transmission path, and a capacitance data sequence after mutual checking is obtained.

5. The method of claim 1, wherein, The capacitance data sequence after mutual checking is segmented and converted into stress values, a preliminary stress data sequence is obtained, the preliminary stress data sequence is converted back into capacitance data, and a reconstructed capacitance data sequence is obtained, including: The capacitance variation trend of the measuring points in the capacitance data sequence after mutual checking is analyzed, the capacitance variation amplitudes between adjacent measuring points are calculated, the mutation positions are identified according to the capacitance variation amplitudes, and the data segmentation intervals are determined; The stress deformation characteristics of the measuring points are extracted in the data segmentation intervals, the corresponding relationship between the stress deformation characteristics and the capacitance variation is analyzed, the capacitance-stress conversion coefficients of each data segmentation interval are generated, and a segmented mapping relationship from capacitance to stress is constructed; The segmented mapping relationship is used to convert and calculate the capacitance data sequence after mutual checking, and the preliminary stress data sequence is obtained through the numerical smooth connection at the segmentation boundaries; The capacitance-stress conversion coefficients are reversely configured to establish a reverse calculation relationship from stress to capacitance, the preliminary stress data sequence is reversely converted, and a reconstructed capacitance data sequence is obtained.

6. The method of claim 1, wherein, The residual errors between the reconstructed capacitance data sequence and the capacitance data sequence after mutual checking are calculated, the spatial gradients of the residual errors along the measuring point positions are extracted, the mutation measuring points are identified, and the mutation measuring points are corrected by using the conversion relationship of adjacent measuring points, and an optimized cross arm stress data sequence is obtained, including: The numerical differences between the reconstructed capacitance data sequence and the capacitance data sequence after mutual checking at the measuring point positions are calculated, measuring point residual data is generated, and the residual variation amount between the measuring points is calculated according to the measuring point residual data; The measuring point residual data and the residual variation amount are used to construct a residual distribution function, the spatial directional derivative of the residual distribution function at the measuring point positions is calculated, and a residual spatial gradient value is obtained; The mutation measuring points are determined by analyzing the measuring point positions whose residual spatial gradient values exceed a preset gradient range, the capacitance variation amounts of the measuring points on both sides of the mutation measuring points are extracted, and a measuring point conversion function is generated. The measurement point conversion function is applied to stress data at the mutation measurement point for correction compensation, to obtain an optimized cross arm stress data sequence.

7. The method of claim 1, wherein, Based on the optimized cross arm stress data, the load bearing state of the cross arm is evaluated, stress abnormal regions are identified, and warning information is generated, including: The stress distribution characteristics of the measurement points in the optimized cross arm stress data sequence are analyzed, the stress change trend of the measurement points is extracted, and a cross arm stress distribution map is constructed. According to the cross arm stress distribution map, the stress concentration degree of each region of the cross arm is calculated, the stress state of the cross arm is evaluated, and a cross arm load bearing state evaluation result is generated. Based on the cross arm load bearing state evaluation result, stress overrun regions are identified, the stress aggregation degree in the stress overrun regions is calculated, and a stress risk assessment index is constructed. The stress risk assessment index is combined with the stress aggregation degree for hierarchical judgment, and warning information containing risk level and regional positioning is generated.

8. A cross arm capacitive stress data accuracy optimization system for implementing the method of any one of claims 1-7, characterized by, The system includes: A data acquisition and temperature compensation unit is configured to acquire a capacitance change signal of the cross arm through a capacitive sensor to obtain an original capacitance data sequence, and perform temperature compensation processing on the original capacitance data sequence to obtain a temperature-compensated capacitance data sequence. A multi-measurement point mutual verification unit is configured to establish a multi-measurement point mutual verification mechanism based on the temperature-compensated capacitance data sequence, select multiple measurement points on the cross arm, and construct a constraint relationship between the capacitance data of each measurement point according to the mechanical balance condition of the cross arm to calculate a theoretical capacitance value of each measurement point. A data compensation unit is configured to compensate the measurement point data by taking the deviation between the actual capacitance value and the theoretical capacitance value as a correction amount to obtain a mutual verification capacitance data sequence. A data conversion unit is configured to convert the mutual verification capacitance data sequence into stress values in segments to obtain a preliminary stress data sequence, and convert the preliminary stress data sequence back into capacitance data to obtain a reconstructed capacitance data sequence. A mutation measurement point correction unit is configured to calculate the residual error between the reconstructed capacitance data sequence and the mutual verification capacitance data sequence, extract the spatial gradient of the residual error along the measurement point position, identify a mutation measurement point, correct the mutation measurement point using the conversion relationship of adjacent measurement points, and obtain an optimized cross arm stress data sequence. A warning unit is configured to evaluate the load bearing state of the cross arm based on the optimized cross arm stress data, identify stress abnormal regions, and generate warning information.

Images