CN-121997122-A - Power grounding state monitoring method and system

CN121997122ACN 121997122 ACN121997122 ACN 121997122ACN-121997122-A

Abstract

The invention discloses a power grounding state monitoring method and a power grounding state monitoring system, which are applied to the field of power equipment state monitoring and comprise the steps of synchronously acquiring a time sequence mechanical state signal and a time sequence electrical state signal of a grounding terminal, extracting and constructing a multidimensional feature vector containing static and dynamic feature components, carrying out differential evaluation and integration on the feature components by utilizing a pre-constructed hybrid analysis model to generate an initial monitoring result, introducing environmental parameters to dynamically calibrate the initial result, and obtaining a calibrated grounding state monitoring result. According to the power grounding state monitoring method provided by the invention, the comprehensive and accurate sensing of the grounding terminal connection state is realized by fusing the mechanical fastening degree and the electrical connectivity information and adopting the intelligent model to carry out collaborative analysis and environment self-adaptive calibration.

Inventors

QIAN TAO
LIU JINLIN
CAI HAIHUI
JIANG SHIJIE
LUO MAOSEN
WANG JINBAO
SONG JIAHUI
LIN ZENGXIAN
ZHANG JIACONG
HE TAO
DONG PENG

Assignees

国网浙江省电力有限公司建德市供电公司
国网浙江省电力有限公司杭州供电公司

Dates

Publication Date: 20260508
Application Date: 20251231

Claims (10)

1. A method for monitoring a power ground state, comprising: acquiring a time sequence mechanical state signal and a time sequence electrical state signal of a target grounding terminal; Integrating the characteristic extraction result of the time sequence mechanical state signal and the characteristic extraction result of the time sequence electrical state signal, and obtaining a multidimensional characteristic vector for representing the comprehensive operation state of the target grounding terminal based on the integration result, wherein the multidimensional characteristic vector at least comprises a static characteristic component and a dynamic characteristic component; Inputting the multi-dimensional characteristic sequence into a pre-constructed grounding state analysis model to obtain a first state evaluation vector corresponding to the static characteristic component and a second state evaluation vector corresponding to the dynamic characteristic component; Generating an initial state monitoring result based on the first state evaluation vector and the second state evaluation vector; And acquiring at least one environmental parameter of the environment where the target grounding terminal is located, and dynamically calibrating the initial state monitoring result based on the environmental parameter to obtain a calibrated grounding state monitoring result.
2. The power-ground-state monitoring method of claim 1, wherein the integrating the feature extraction result of the time-series mechanical-state signal and the feature extraction result of the time-series electrical-state signal and obtaining a multidimensional feature vector for characterizing the integrated operation state of the target ground terminal based on the integration result comprises: performing time domain analysis on the time sequence mechanical state signal, extracting pressure change frequency as a dynamic characteristic component, and extracting pressure amplitude as a static characteristic component; performing time domain analysis on the time sequence electrical state signal, calculating the fluctuation range of the grounding resistance in the power loop as a dynamic characteristic component, and calculating the standard deviation of the grounding resistance as a static characteristic component; And synchronously aligning and combining the extracted pressure change frequency, the pressure amplitude, the fluctuation range and the standard deviation according to corresponding time stamps to construct the multidimensional feature vector.
3. The method of claim 1, wherein the ground state analysis model is a hybrid model, the inputting the multi-dimensional feature sequence into a pre-built ground state analysis model to obtain a first state evaluation vector corresponding to the static feature component and a second state evaluation vector corresponding to the dynamic feature component, comprising: Inputting the static characteristic component into a random forest model for static evaluation, and outputting the first state evaluation vector; And inputting the dynamic characteristic component into a long-term and short-term memory network model for dynamic evaluation, and outputting the second state evaluation vector.
4. The method of power ground state monitoring of claim 1, wherein generating an initial state monitoring result based on the first state evaluation vector and the second state evaluation vector comprises: performing standardization processing on the second state evaluation vector, and converting the second state evaluation vector into a probability vector with the same dimension as the first state evaluation vector; based on an attention mechanism, dynamically generating a fusion weight between the first state evaluation vector and the probability vector according to the multidimensional feature vector of the current time step, and carrying out weighted fusion on the first state evaluation vector and the probability vector according to the fusion weight to obtain comprehensive state probability distribution; and outputting the state category identification and the corresponding state probability of the target grounding terminal according to the comprehensive state probability distribution, and generating the initial state monitoring result.
5. The method for monitoring the ground state of power according to claim 1, wherein dynamically calibrating the initial state monitoring result based on the environmental parameter to obtain a calibrated ground state monitoring result comprises: Constructing a fuzzy evaluation system, and mapping the initial state monitoring result into a corresponding membership expression based on the fuzzy evaluation system; Dynamically adjusting the weight of the fuzzy evaluation system based on the environmental parameter, and processing the membership grade identifier based on the adjusted fuzzy evaluation system to generate a calibration factor; and correcting the initial state monitoring result by using the calibration factor to obtain the calibrated grounding state monitoring result.
6. A power ground condition monitoring system, comprising: the acquisition module is used for acquiring a time sequence mechanical state signal and a time sequence electrical state signal of the target grounding terminal; The integration module is used for integrating the characteristic extraction result of the time sequence mechanical state signal and the characteristic extraction result of the time sequence electrical state signal, and obtaining a multidimensional characteristic vector for representing the comprehensive operation state of the target grounding terminal based on the integration result, wherein the multidimensional characteristic vector at least comprises a static characteristic component and a dynamic characteristic component; The evaluation module is used for inputting the multidimensional characteristic sequence into a pre-constructed grounding state analysis model to obtain a first state evaluation vector corresponding to the static characteristic component and a second state evaluation vector corresponding to the dynamic characteristic component; The generating module is used for generating an initial state monitoring result based on the first state evaluation vector and the second state evaluation vector; And the calibration module is used for acquiring at least one environmental parameter of the environment where the target grounding terminal is positioned, and dynamically calibrating the initial state monitoring result based on the environmental parameter to obtain a calibrated grounding state monitoring result.
7. The power ground condition monitoring system of claim 6, wherein said integration module comprises: The first time domain analysis unit is used for performing time domain analysis on the time sequence mechanical state signal, extracting the pressure change frequency as a dynamic characteristic component and extracting the pressure amplitude as a static characteristic component; The second time domain analysis unit is used for performing time domain analysis on the time sequence electrical state signals, calculating the fluctuation range of the grounding resistance in the power loop as a dynamic characteristic component, and calculating the standard deviation of the grounding resistance as a static characteristic component; and the alignment unit is used for synchronously aligning and combining the extracted pressure change frequency, the pressure amplitude, the fluctuation range and the standard deviation according to the corresponding time stamps to construct the multidimensional feature vector.
8. The power ground state monitoring system of claim 6, wherein the ground state analysis model is a hybrid model, the evaluation module comprising: The static evaluation unit is used for inputting the static characteristic components into a random forest model for static evaluation and outputting the first state evaluation vector; And the dynamic evaluation unit is used for inputting the dynamic characteristic component into a long-short-term memory network model for dynamic evaluation and outputting the second state evaluation vector.
9. The power ground state monitoring system of claim 6, wherein the generation module comprises: the normalization processing unit is used for performing normalization processing on the second state evaluation vector and converting the second state evaluation vector into a probability vector with the same dimension as the first state evaluation vector; The weighted fusion unit is used for dynamically generating fusion weights between the first state evaluation vector and the probability vector according to the multidimensional feature vector of the current time step based on an attention mechanism, and carrying out weighted fusion on the first state evaluation vector and the probability vector according to the fusion weights to obtain comprehensive state probability distribution; and the probability generation unit is used for outputting the state category identification of the target grounding terminal and the corresponding state probability according to the comprehensive state probability distribution, and generating the initial state monitoring result.
10. The power ground condition monitoring system of claim 6, wherein the calibration module comprises: The system construction unit is used for constructing a fuzzy evaluation system and mapping the initial state monitoring result into a corresponding membership expression based on the fuzzy evaluation system; The dynamic adjustment unit is used for dynamically adjusting the weight of the fuzzy evaluation system based on the environmental parameter, and processing the membership grade identification based on the adjusted fuzzy evaluation system to generate a calibration factor; And the result calibration unit is used for correcting the initial state monitoring result by using the calibration factor to obtain the calibrated grounding state monitoring result.

Description

Power grounding state monitoring method and system Technical Field The invention relates to the technical field of power equipment state monitoring, in particular to a power grounding state monitoring method and system. Background The grounding terminal is used as an important safety component of the power equipment, the compaction force of the grounding terminal directly influences the grounding effect, the power equipment is often operated outdoors or in severe industrial environments, the grounding terminal is possibly loosened due to factors such as temperature change and vibration, if the nut of the grounding terminal is loosened to cause poor grounding, the current cannot be effectively led into the ground when the equipment fails, and serious consequences such as safety of the equipment and even fire disaster are threatened. In the prior art, a monitoring method for a grounding terminal mainly relies on a mechanical connection signal to push out a grounding state, and the single mechanical signal-dependent evaluation model is difficult to comprehensively and accurately reflect a complicated grounding connection degradation process under the combined action of various factors such as electric corrosion, environmental change and the like. In particular, for the electrical performance degradation caused by non-mechanical factors such as corrosion and pollution of the contact surface, the method has obvious monitoring blind areas, and the diagnosis result lacks sufficient correlation and accuracy. The method directly causes insufficient diagnosis accuracy of the grounding state under the complex working condition, has higher false alarm and missing report rate, and is difficult to provide reliable basis for preventive maintenance. Disclosure of Invention The invention provides a power grounding state monitoring method and a system, which are used for solving the technical problem that complicated degradation modes of a grounding terminal are difficult to comprehensively and accurately identify due to single monitoring dimension, so as to realize the effect of cooperative sensing and comprehensive evaluation of mechanical connection and electrical performance states of the grounding terminal. In order to solve the above technical problems, an embodiment of the present invention provides a method for monitoring a power ground state, including: acquiring a time sequence mechanical state signal and a time sequence electrical state signal of a target grounding terminal; Integrating the characteristic extraction result of the time sequence mechanical state signal and the characteristic extraction result of the time sequence electrical state signal, and obtaining a multidimensional characteristic vector for representing the comprehensive operation state of the target grounding terminal based on the integration result, wherein the multidimensional characteristic vector at least comprises a static characteristic component and a dynamic characteristic component; Inputting the multi-dimensional characteristic sequence into a pre-constructed grounding state analysis model to obtain a first state evaluation vector corresponding to the static characteristic component and a second state evaluation vector corresponding to the dynamic characteristic component; Generating an initial state monitoring result based on the first state evaluation vector and the second state evaluation vector; And acquiring at least one environmental parameter of the environment where the target grounding terminal is located, and dynamically calibrating the initial state monitoring result based on the environmental parameter to obtain a calibrated grounding state monitoring result. As one preferable aspect, the integrating the feature extraction result of the time-series mechanical state signal and the feature extraction result of the time-series electrical state signal, and obtaining a multidimensional feature vector for characterizing the comprehensive operation state of the target ground terminal based on the integration result, includes: performing time domain analysis on the time sequence mechanical state signal, extracting pressure change frequency as a dynamic characteristic component, and extracting pressure amplitude as a static characteristic component; performing time domain analysis on the time sequence electrical state signal, calculating the fluctuation range of the grounding resistance in the power loop as a dynamic characteristic component, and calculating the standard deviation of the grounding resistance as a static characteristic component; And synchronously aligning and combining the extracted pressure change frequency, the pressure amplitude, the fluctuation range and the standard deviation according to corresponding time stamps to construct the multidimensional feature vector. As one preferable mode, the ground state analysis model is a hybrid model, and the step of inputting the multi-dimensional feature sequence into a p