A method for dynamically predicting grounding resistance of a power transmission tower

Abstract

The application is a kind of power transmission tower grounding resistance dynamic prediction method, belonging to the technical field of power transmission line, including data preprocessing of collected power transmission tower grounding test data, eliminating interference data; according to soil environment and grounding body aging degree, establishing power transmission tower grounding resistance correction formula; considering the inhibition of seasonal resistance buried depth to seasonal fluctuation in seasonal alpine region, constructing seasonal coefficient, and carrying out soil resistivity conversion in permafrost region; constructing multivariate nonlinear regression model, solving by least square method, and obtaining power transmission tower grounding resistance prediction value. According to the measured data set, the application screens the core influence parameters, establishes a ternary quadratic multivariate nonlinear regression model, effectively fits the complex nonlinear relationship between the grounding resistance and the influence factors, realizes the efficient and accurate dynamic prediction of the grounding resistance, and provides technical support for lightning protection evaluation and state maintenance of power transmission line.

Description

Technical Field

[0001] This invention belongs to the field of power transmission line technology, specifically relating to a method for dynamic prediction of grounding resistance of power transmission towers. Background Technology

[0002] Transmission lines are the main arteries of the power grid, forming a large-scale system distributed across the ground, with some reaching thousands of kilometers in geometric scale. As my country's power grid construction continues to expand, the number of overhead transmission lines is increasing, and line corridors are becoming denser, placing increasingly stringent requirements on tower grounding facilities. The grounding resistance of transmission towers is a key factor affecting the safety and reliability of line operation, directly impacting personal and equipment safety. High grounding resistance easily triggers lightning backflashovers, leading to line tripping, which is particularly prominent in areas with frequent lightning strikes, high soil resistivity, and complex terrain. Lightning tripping can cause equipment damage, line outages, and even large-scale power outages, resulting in significant economic losses and social impact. Therefore, conducting safety assessment research on the grounding resistance of transmission towers is of great significance for ensuring the stable operation of the power system.

[0003] Tower grounding resistance is divided into power frequency grounding resistance and impulse grounding resistance, which differ significantly. Assessing lightning protection safety solely based on power frequency grounding resistance will lead to inaccurate safety judgments under lightning strikes; therefore, accurately understanding the characteristics of impulse grounding resistance is crucial. However, due to the high amplitude and high frequency characteristics of lightning, on-site measurement of impulse grounding resistance is extremely difficult, and existing simulation methods suffer from low accuracy in soil ionization calculations and difficulty in converging nonlinear processes.

[0004] Traditional power frequency grounding resistance measurement relies on manual on-site operation, which requires power outages, resulting in low efficiency, high cost, and significant operational risks. Furthermore, it can only obtain static values ​​at a specific moment, failing to reflect the dynamic changes in grounding resistance over time and environmental factors. Existing data-driven prediction methods often consider single factors, lacking generalization ability and failing to meet engineering accuracy requirements. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a dynamic prediction method for the grounding resistance of transmission towers. Based on the measured dataset, core influencing parameters are selected, and a ternary quadratic multivariate nonlinear regression model is established to effectively fit the complex nonlinear relationship between grounding resistance and influencing factors, thereby achieving efficient and accurate dynamic prediction of grounding resistance and providing technical support for lightning protection assessment and condition-based maintenance of transmission lines.

[0006] To solve the above-mentioned technical problems, the present invention provides a method for dynamically predicting the grounding resistance of transmission towers, comprising the following steps, which are performed sequentially: Step 1: Perform data preprocessing on the collected transmission tower grounding test data to eliminate interference data; Step 2: Based on the soil environment and the aging degree of the grounding electrode, establish a correction formula for the grounding resistance of the transmission tower; Step 3: Considering the suppression of seasonal fluctuations in grounding resistance burial depth in cold regions, construct a seasonal coefficient and perform soil resistivity conversion in permafrost areas. Step 4: Construct a multivariate nonlinear regression model and solve it using the least squares method to obtain the predicted value of the grounding resistance of the transmission tower.

[0007] The data preprocessing described in step one includes identifying and removing duplicate data values, correcting and classifying abnormal data values, and marking and isolating missing data values.

[0008] The correction formula for the grounding resistance of the transmission tower mentioned in step two is as follows: , In the formula: r eff The effective radius of the grounding electrode. d For the burial depth of the grounding electrode, t For the service life of the grounding electrode, f corr As a corrosion oxide layer correction factor, f pH As a soil pH correction factor, r For soil resistivity, L The length of the grounding electrode. r The original radius of the grounding electrode. h This is the correction value for the equivalent depth of the grounding electrode; The effective radius of the grounding electrode is restored as follows: , , , In the formula: Δr This is the amount of thinning due to corrosion of the grounding electrode radius. K The soil corrosion rate coefficient, r eff The effective radius of the grounding electrode is in millimeters (mm). y For the year, K The unit is mm / y; Corrosion oxide layer correction factor: , In the formula: f corr This is a corrosion oxide layer resistance correction factor; the oxide layer on the grounding electrode surface leads to an increase in contact resistance.

[0009] The seasonal coefficient mentioned in step three is defined as the ratio of the annual maximum soil resistivity to the measured soil resistivity for the current month. , Convert the measured values ​​for any month to the standard values ​​under frozen soil conditions: , In the formula: r max The annual maximum soil resistivity, The measured soil resistivity for that month. Standard values ​​for frozen soil working conditions For the first i Monthly measured soil resistivity.

[0010] The multivariate nonlinear regression model described in step four is as follows: , In the formula: R pred The predicted grounding resistance value (Ω) is given by the model. r The measured soil resistivity (Ω·m); ψ This refers to the stratified monthly seasonal coefficients corresponding to the month and burial depth; t The service life (in years) of the tower grounding electrode; β 0 For constant terms; β 1 , β 2 , β 3 The coefficient of the linear term; β 4 , β 5 , β 6 The coefficient of the quadratic term; β 7 , β 8 , β 9 The coefficients of the interaction terms; β 1 - β 9 The method of least squares was used to perform multivariate nonlinear regression analysis on the field measured data, and the objective function was to minimize the sum of squared residuals between the measured grounding resistance and the calculated value of the model.

[0011] The beneficial effects of this invention are as follows: This invention provides a dynamic prediction method for the grounding resistance of transmission towers. The model's independent variables are entirely derived based on the Dwight formula and multi-dimensional correction theory, with each parameter having a clear physical meaning, thus avoiding the interpretability issues of black-box models. By introducing layered and monthly seasonal coefficients and nonlinear terms, the problem of large calculation errors in grounding resistance in seasonally frozen soil areas is solved, with prediction errors controlled within 15%. Only four easily obtainable parameters need to be input: soil resistivity, test month, grounding electrode burial depth (generally 0.8m by default), and operating years. No complex grounding grid topology calculations are required, enabling batch prediction of grounding resistance for large-scale transmission towers. Attached Figure Description

[0012] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0013] Figure 1 Table of measured seasonal coefficients of soil at different burial depths in specific embodiments of the present invention; Figure 2 This is a schematic diagram illustrating the seasonal coefficient variation trend in a specific embodiment of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] In a specific embodiment, using the original tower grounding test data collected by the Jilin Provincial Electric Power Research Institute in the Changchun area as the basic data, invalid data interference is eliminated through multi-dimensional verification and cleaning. This process includes the following steps: Duplicate value identification and removal: A multi-field combination uniqueness verification mechanism is constructed, with test records, measured grounding resistance, calculated resistance value, and grounding electrode form as the core verification fields. It automatically identifies and removes duplicate invalid test records that are completely identical in all fields, eliminates data redundancy caused by duplicate sampling or data entry errors, and ensures the uniqueness and accuracy of the training dataset.

[0016] Outlier Correction and Classification: A tiered processing strategy is adopted to address data anomalies. For extreme outliers exceeding the reasonable range for tower grounding resistance, the qualified resistance values ​​from the same batch of tests conducted during the same period are used to replace them, ensuring data authenticity while avoiding interference from extreme values ​​with model fitting. For unqualified data where the calculated resistance exceeds the corresponding design threshold, the causes of the anomalies are traced and marked one by one. Typical causes include insufficient grounding electrode length, grounding point being covered by concrete, and ineffective connection of the grounding down conductor, providing support for subsequent grounding hazard analysis.

[0017] Missing value marking and isolation: Test records with missing key fields are uniformly marked as "not valid test" and archived separately. Such records are not included in the statistical analysis of grounding resistance compliance and the regression model training dataset, effectively eliminating the interference of invalid data on the accuracy of statistical results and the model training effect.

[0018] The theoretical calculation of grounding resistance is based on the classic Dwight formula (1936). The formula for calculating the foundation grounding resistance of a single vertical grounding electrode is as follows:

[0019] In the formula: r Soil resistivity (Ω·m). L The length of the grounding electrode is in meters (m). r The original radius of the grounding electrode (m).

[0020] Transmission towers commonly employ composite grounding grids consisting of multiple vertical grounding electrodes and horizontal connecting strips, with the most typical forms being ring-type and radial grounding grids. For a grounding grid composed of n vertical grounding electrodes of the same specification, its total power frequency grounding resistance can be calculated by multiplying the resistance of a single electrode by the grounding electrode utilization factor. or The formula is then expanded to:

[0021] In the formula: or The grounding electrode utilization factor (valued at 0.65-0.85) is used to correct for the current resistance shielding effect between multiple grounding electrodes. For the conventional ring arrangement of four vertical grounding electrodes on transmission towers, the utilization factor is typically 0.75-0.8; for a radial arrangement, it is typically 0.7-0.75. The grounding resistance of the horizontal connection strip usually accounts for less than 15%, and its variation pattern is completely consistent with that of the vertical grounding electrodes. Therefore, its influence can be included in the utilization factor in engineering calculations without separate calculation.

[0022] Considering the long-term effects of soil environment and aging of grounding electrodes, the actual grounding resistance in operation needs to be corrected in multiple dimensions. The complete correction formula is as follows:

[0023] In the formula: r eff The effective radius of the grounding electrode. d For the burial depth of the grounding electrode, t For the service life of the grounding electrode, f corr As a corrosion oxide layer correction factor, f pH The soil pH correction factor is used. Based on the above theoretical formula, this patent systematically screens various potential influencing factors such as tower operating environment, equipment status, and time dimension. Combining engineering experience and statistical analysis results, non-critical influencing factors are eliminated, and soil resistivity and operating years are finally determined as the two basic core independent variables of the model.

[0024] Soil resistivity (core intrinsic independent variable): Soil resistivity is the most fundamental intrinsic factor determining grounding resistance, appearing directly in the numerator of Dwight's basic formula. Measured data show a significant positive correlation (correlation coefficient > 0.7) between soil resistivity and calculated grounding resistance, covering the typical soil resistivity range of 100–500 Ω·m in the Changchun area, making it a fundamental input parameter for the model. Existing formulas for correcting soil resistivity include: Moisture content correction:

[0025] Burial depth correction:

[0026] However, the above-mentioned sub-item corrections require real-time collection of on-site parameters such as soil moisture content and temperature, which has poor engineering practicality and does not take into account the freeze-thaw cycle effect in seasonally frozen soil areas. Therefore, this invention provides a comprehensive correction of soil resistivity in seasonally frozen soil areas and determination of stratified monthly seasonal coefficients.

[0027] Service life (core independent variable for corrosion and aging): Service life is the only quantifiable and readily available parameter characterizing the long-term corrosion and aging effect of a grounding electrode. Its impact on grounding resistance is achieved through two mechanisms: thinning of the grounding electrode due to corrosion and increase in the resistance of the surface oxide layer. The corresponding theoretical correction formula is as follows: Grounding electrode corrosion thinning correction:

[0028]

[0029]

[0030] In the formula: Δr This is the amount of thinning due to corrosion of the grounding electrode radius.K The soil corrosion rate coefficient, r eff The effective radius of the grounding electrode (the minimum effective radius is specified to be no less than 10% of the original radius).

[0031] Corrosion and oxide layer correction:

[0032] In the formula: f corr This is a corrosion oxide layer resistance correction factor, reflecting the increase in contact resistance caused by the oxide layer on the grounding electrode surface.

[0033] The above formula shows that the grounding resistance increases non-linearly with the service life. This trend has been verified by the measured data of this patent: the grounding failure rate of towers with a service life of >10 years is 42% higher than that of newly built towers. Existing data-driven models often ignore the non-linear effect of corrosion, resulting in a significant decrease in long-term prediction accuracy. This patent uses the service life as the core independent variable, effectively solving this problem.

[0034] Existing technologies generally use a single fixed seasonal coefficient for soil resistivity conversion, failing to consider the suppressive effect of grounding electrode burial depth on seasonal fluctuations. This can lead to calculation errors of over ±30% in the seasonally frozen soil region of Northeast China, severely impacting the accuracy of prediction models. To address this core deficiency, this embodiment comprehensively corrects soil resistivity based on measured data, quantifying the coupled effects of multiple factors such as moisture content, temperature, freeze-thaw cycles, and burial depth into layered and monthly seasonal coefficients. ψ This enables rapid and accurate engineering-based correction. The seasonal coefficient... ψ Defined as: the ratio of the annual maximum soil resistivity to the measured soil resistivity for the current month, i.e.:

[0035] Its physical essence is a comprehensive normalization correction for all seasonal influencing factors of soil resistivity, which can directly convert the measured value of any month into the standard value under the worst working conditions in the design year:

[0036] This method is based on five consecutive years of in-situ measured soil resistivity data from the Jilin Electric Power Research Institute in Changchun area (including typical silty clay and black soil distribution areas such as Beihu Science and Technology Development Zone, with a maximum seasonal frost depth of 1.7-1.8m) from 2020 to 2025. Combined with DL / T 475-2017 "Guidelines for Measurement of Characteristic Parameters of Grounding Devices" and DL / T 5091-2022 "Design Specifications for 110kV~750kV Overhead Transmission Lines", the method determines the monthly grounding seasonal coefficients for Changchun area according to the burial depth of the grounding electrode. Specific values ​​are as follows: Figure 1As shown, the seasonal coefficient variation trend is as follows: Figure 2 As shown.

[0037] Compared with existing theoretical component correction methods and single fixed coefficient methods, the stratified monthly seasonal coefficients of this invention have the following advantages: High comprehensive correction accuracy: It incorporates the coupled effects of multiple factors such as moisture content, temperature, freeze-thaw cycles, and burial depth. The measured correction error in Changchun area is ≤5%, which is far superior to the traditional single coefficient method. Highly practical for engineering applications: No additional on-site parameter collection is required; correction coefficients can be obtained by looking up tables based on the test month and the burial depth of the grounding electrode, supporting batch calculation and dynamic prediction of grounding resistance of large-scale towers; Adapted to the characteristics of permafrost regions: It clearly distinguishes the correction rules above and below the permafrost layer, filling the technical gap in the accurate correction of grounding resistance in the seasonal permafrost region of Northeast China.

[0038] Thus, this embodiment has identified three core independent variables: soil resistivity, seasonal coefficient, and years of operation. These three variables respectively cover the three dimensions of soil intrinsic characteristics, short-term environmental changes, and long-term equipment aging. Furthermore, the data is readily available and the coupling relationship is clear, which can effectively fit the complex nonlinear variation law of grounding resistance, laying a solid foundation for the subsequent construction of a ternary quadratic multivariate nonlinear regression model.

[0039] Based on the three core independent variables—soil resistivity, seasonal coefficient, and service life—determined by the aforementioned methods, this invention constructs a ternary quadratic multivariate nonlinear regression model for the dynamic prediction of grounding resistance of transmission towers. This model retains the physical meaning of the theoretical formulas while correcting systematic errors in theoretical calculations through fitting measured data. Simultaneously, it avoids the input requirements of complex grounding grid topology parameters, achieving a balance between accuracy and engineering practicality.

[0040] Model selection criteria: There is a significant nonlinear coupling relationship between grounding resistance and various influencing factors: Soil resistivity and grounding resistance have an approximately linear relationship, but are subject to nonlinear modulation by seasonal coefficients; The service life exhibits a power-law nonlinear relationship with the grounding resistance due to corrosion thinning and oxide layer effects. There are interactions among the various factors; for example, the corrosion rate is faster in areas with high soil resistivity, and seasonal fluctuations have a more significant impact on grounding resistance.

[0041] Traditional linear regression models cannot capture the aforementioned nonlinear relationships, while black-box models such as neural networks have poor interpretability and require large amounts of training data, making them difficult to pass engineering audits in the power industry. The ternary quadratic multivariate nonlinear regression model can accurately describe the quadratic nonlinear relationships and interactions between variables, and its simple model structure, clear physical meaning of parameters, and high computational efficiency fully meet the needs of engineering applications.

[0042] The general expression of the ternary quadratic multivariate nonlinear regression model constructed in this invention is as follows:

[0043] In the formula: Rpred is the predicted grounding resistance value (Ω) from the model. ρ is the measured soil resistivity (Ω·m); ψ is the stratified monthly seasonal coefficient corresponding to the month and burial depth; t represents the service life (in years) of the tower grounding electrode; β0 is a constant term; β1, β2, β3 are the coefficients of the linear term; β4, β5, and β6 are the coefficients of the quadratic term; β7, β8, and β9 are the coefficients of the interaction term.

[0044] The model coefficients are solved using the least squares method, minimizing the sum of squared residuals between the predicted and measured values:

[0045] Model training and validation: The Changchun area tower grounding test dataset, which uses the above-mentioned standardized cleaning process, contains 9,214 valid tower test records, covering the two mainstream grounding types of ring and radial grounding, with soil resistivity ranging from 100 to 500 Ω·m, operating years from 0 to 25 years, and testing time covering all 12 months of the year.

[0046] Training set: 70% of the samples are randomly selected and used to solve for the model coefficients; Test set: The remaining 30% of the samples are used to verify the model's generalization ability.

[0047] The model training data covers the mainstream grounding methods, soil types, and operating years in the Changchun area, and can be directly applied to other seasonally frozen soil areas in Northeast China.

[0048] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0049] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for dynamically predicting the grounding resistance of transmission towers, characterized in that: Includes the following steps, And the following steps are performed in sequence: Step 1: Perform data preprocessing on the collected transmission tower grounding test data to eliminate interference data; Step 2: Based on the soil environment and the aging degree of the grounding electrode, establish a correction formula for the grounding resistance of the transmission tower; Step 3: Considering the suppression of seasonal fluctuations in grounding resistance burial depth in cold regions, construct a seasonal coefficient and perform soil resistivity conversion in permafrost areas. Step 4: Construct a multivariate nonlinear regression model and solve it using the least squares method to obtain the predicted value of the grounding resistance of the transmission tower.

2. The method for dynamically predicting the grounding resistance of transmission towers according to claim 1, characterized in that: The data preprocessing described in step one includes identifying and removing duplicate data values, correcting and classifying abnormal data values, and marking and isolating missing data values.

3. The method for dynamically predicting the grounding resistance of transmission towers according to claim 1, characterized in that: The correction formula for the grounding resistance of the transmission tower mentioned in step two is as follows: ， In the formula: r eff The effective radius of the grounding electrode. d For the burial depth of the grounding electrode, t For the service life of the grounding electrode, f corr As a corrosion oxide layer correction factor, f pH As a soil pH correction factor, ρ For soil resistivity, L The length of the grounding electrode. r The original radius of the grounding electrode. h This is the correction value for the equivalent depth of the grounding electrode; The effective radius of the grounding electrode is restored as follows: ， ， ， In the formula: Δr This is the amount of thinning due to corrosion of the grounding electrode radius. K The soil corrosion rate coefficient, r eff The effective radius of the grounding electrode is in millimeters (mm). y For the year, K The unit is mm / y; Corrosion oxide layer correction factor: ， In the formula: f corr This is a corrosion oxide layer resistance correction factor; the oxide layer on the grounding electrode surface leads to an increase in contact resistance.

4. The method for dynamically predicting the grounding resistance of transmission towers according to claim 1, characterized in that: The seasonal coefficient mentioned in step three is defined as the ratio of the annual maximum soil resistivity to the measured soil resistivity for the current month. ， Convert the measured values ​​for any month to the standard values ​​under frozen soil conditions: ， In the formula: ρ max The annual maximum soil resistivity, The measured soil resistivity for that month. Standard values ​​for frozen soil working conditions For the first i Monthly measured soil resistivity.

5. The method for dynamically predicting the grounding resistance of transmission towers according to claim 1, characterized in that: The multivariate nonlinear regression model described in step four is as follows: ， In the formula: R pred The predicted grounding resistance value (Ω) is given by the model. ρ The measured soil resistivity (Ω·m); ψ This refers to the stratified monthly seasonal coefficients corresponding to the month and burial depth; t The service life (in years) of the tower grounding electrode; β 0 For constant terms; β 1 , β 2 , β 3 The coefficient of the linear term; β 4 , β 5 , β 6 The coefficient of the quadratic term; β 7 , β 8 , β 9 The coefficients of the interaction terms; β 1 - β 9 The method of least squares was used to perform multivariate nonlinear regression analysis on the field measured data, and the objective function was to minimize the sum of squared residuals between the measured grounding resistance and the calculated value of the model.

Images