Search

CN-121542910-B - Method and system for monitoring moisture of capacitive transformer bushing

CN121542910BCN 121542910 BCN121542910 BCN 121542910BCN-121542910-B

Abstract

The invention relates to the technical field of electrical equipment state monitoring, and discloses a method and a system for monitoring the moisture of a capacitive transformer bushing. The method comprises the steps of obtaining initial monitoring data flow from monitoring nodes deployed in transformer bushings, performing feature extraction based on time evolution analysis on the data flow to obtain a preliminary humidity response sequence, a preliminary temperature response sequence and a potential disturbance identification sequence, performing correlation verification of humidity and temperature sequences, calculating coupling influence degree of the disturbance identification sequences on the humidity and temperature sequences, constructing a dynamic probability response matrix according to verification results and coupling influence degree, deducing a damping evaluation index, and performing multi-level signal correction on the initial monitoring data flow according to the index to analyze core humidity components and core temperature components reflecting damping progress. The invention can separate the damping signal from the temperature and disturbance effect, and realize the dynamic and accurate assessment of the damping state of the transformer bushing.

Inventors

  • WANG XIAOHUA

Assignees

  • 搏世因（北京）高压电气有限公司

Dates

Publication Date
20260508
Application Date
20260116

Claims (9)

  1. 1. The method for monitoring the moisture of the bushing of the capacitive transformer is characterized by comprising the following steps of: acquiring an initial monitoring data stream from a monitoring node deployed in a transformer bushing; executing a feature extraction strategy based on time evolution analysis on the initial monitoring data stream to obtain a preliminary humidity response sequence, a preliminary temperature response sequence and a potential disturbance identification sequence; performing relevance verification of the preliminary humidity response sequence and the preliminary temperature response sequence, and calculating coupling influence degree of the potential disturbance identification sequence on the preliminary humidity response sequence and the preliminary temperature response sequence; constructing a dynamic probability response matrix according to the correlation verification result and the coupling influence degree, and deducing a damping evaluation index based on the dynamic probability response matrix; Performing multi-level signal correction on the initial monitoring data stream according to the damping evaluation index, so as to analyze a core humidity component and a core temperature component reflecting the damping process; The feature extraction strategy based on time evolution analysis is performed on the initial monitoring data stream, and comprises the following steps: constructing a sliding evaluation window for the initial monitoring data stream; identifying and dividing a data segment with continuous stationary attribute and a data segment with transient non-stationary attribute in the initial monitoring data stream in the sliding evaluation window; For the data segment with continuous and stable attribute, adopting a trend stability detection method to separate a baseline humidity component and a baseline temperature component, wherein the baseline humidity component and the baseline temperature component form a preliminary humidity response sequence and partial basic data of the preliminary temperature response sequence; And starting an abnormal mode capturing method for the data segment with transient non-stationary attribute, generating an initial disturbance label with a time mark, and forming a potential disturbance identification sequence by a plurality of sets of the initial disturbance labels.
  2. 2. The method for monitoring the moisture content of the bushing of the capacitive transformer according to claim 1, wherein the method for capturing the abnormal mode is started to generate an initial disturbance tag with a time mark, and the method comprises the following steps of: calculating the local energy entropy and the neighborhood variance of each sampling point in the data segment with transient non-stationary attribute; Inputting the local energy entropy and the neighborhood variance into a preset disturbance classification decision tree; and outputting an initial disturbance type code and initial confidence coefficient through the disturbance classification decision tree, and binding the initial disturbance type code, the initial confidence coefficient and a corresponding sampling point time stamp to form an initial disturbance label.
  3. 3. A method of monitoring moisture in a capacitive transformer bushing according to claim 2, characterized in that said performing a correlation verification of said preliminary humidity response sequence with said preliminary temperature response sequence comprises the steps of: comparing the change rate mode of the preliminary humidity response sequence with the change rate mode of the preliminary temperature response sequence under the same time domain frame; When the change rate mode of the preliminary humidity response sequence and the change rate mode of the preliminary temperature response sequence are in the same-direction covariant relation, calculating covariant intensity coefficient and phase delay amount, and taking the covariant intensity coefficient and the phase delay amount as forward association characteristics; When the change rate mode of the preliminary humidity response sequence and the change rate mode of the preliminary temperature response sequence are in a reverse change relation, recording the occurrence frequency and the change amplitude of a reverse change event, and taking the occurrence frequency and the change amplitude of the reverse change event as abnormal correlation characteristics.
  4. 4. A method of monitoring a capacitive transformer bushing according to claim 3, wherein said calculating the degree of coupling influence of said potential disturbance identification sequence on said preliminary humidity response sequence and said preliminary temperature response sequence comprises the steps of: positioning a corresponding disturbed data interval in the preliminary humidity response sequence and the preliminary temperature response sequence according to the time stamp in the initial disturbance label; Respectively calculating the data distribution distortion degree of the disturbed data interval in the preliminary humidity response sequence and the data distribution distortion degree of the disturbed data interval in the preliminary temperature response sequence; Fusing the data distribution distortion degree of the preliminary humidity response sequence, the data distribution distortion degree of the preliminary temperature response sequence and the initial confidence degree in the initial disturbance label, and calculating the local coupling influence factor of the single initial disturbance label through a weighted evaluation model; and summarizing local coupling influence factors of all the initial disturbance tags, and generating global coupling influence degree by combining timestamp density analysis based on a spatial neighborhood relation.
  5. 5. The method of claim 4, wherein said constructing a dynamic probability response matrix comprises the steps of: taking the forward association features and the abnormal association features as row vector basic elements of a matrix; Taking the global coupling influence degree as a column vector adjustment coefficient of a matrix; And carrying out fusion operation on the row vector basic elements and the column vector adjustment coefficients through a matrix transformation rule, and constructing a dynamic probability response matrix with elements being the occurrence probability of the damp state.
  6. 6. The method of claim 5, further comprising, prior to said constructing a dynamic probability response matrix based on said correlation verification result and said coupling influence, the steps of: Checking whether an initial disturbance label group with the same initial disturbance type code and continuous time stamps exists in the potential disturbance identification sequence; If the initial disturbance label group exists, disturbance event merging and confidence strengthening operation is carried out on the initial disturbance label group to form a composite disturbance mark, and the potential disturbance mark sequence is updated by using the composite disturbance mark.
  7. 7. The method of claim 6, wherein the performing disturbance event merging and confidence enhancing operations on the initial disturbance tag group to form a composite disturbance flag comprises the steps of: Extracting time stamps of all initial disturbance tags in the initial disturbance tag group, and determining a composite time interval by taking the earliest time stamp as a starting point and the latest time stamp as an ending point; calculating a weighted average value of initial confidence coefficients of all initial disturbance tags in the initial disturbance tag group, and multiplying the weighted average value by a preset strengthening coefficient to obtain a strengthened confidence coefficient; And packaging the initial disturbance type code, the composite time interval and the reinforced confidence coefficient to generate a composite disturbance mark.
  8. 8. The method of claim 7, wherein said performing multi-level signal correction on said initial monitored data stream based on said moisture assessment indicator comprises the steps of: A first-stage correction step of adaptively setting a plurality of signal amplitude filtering thresholds in the initial monitoring data stream according to the damp evaluation index; The second-stage correction is carried out, namely the data flow after the first-stage correction is compared with the composite time interval of the composite disturbance mark, the data points falling in the composite time interval are removed, and the removed data gaps are subjected to context-based smooth interpolation; And third-stage correction, namely inputting the data stream subjected to the second-stage correction processing into an optimized filter taking the dynamic probability response matrix as a parameter, and finally separating a core humidity component and a core temperature component.
  9. 9. A capacitive transformer bushing moisture monitoring system comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that the processor, when executing the computer program, realizes the steps of the capacitive transformer bushing moisture monitoring method according to any one of claims 1 to 8.

Description

Method and system for monitoring moisture of capacitive transformer bushing Technical Field The invention relates to the technical field of state monitoring of electrical equipment, in particular to a method and a system for monitoring the moisture of a capacitive transformer bushing. Background The bushing of the capacitive transformer is an important insulating component of the power transformer, and the internal insulation of the bushing is wetted, so that the electrical performance is directly reduced, and even serious accidents are caused. The existing monitoring technology mainly relies on collecting single or few parameters such as dielectric loss factor, equivalent capacitance or end screen current of the sleeve, and judges whether the sleeve is abnormal or not by setting a fixed threshold or performing simple trend comparison. Such methods treat the monitored data as static or quasi-static signals, the analysis of which is premised on parameter changes mainly caused by insulation state changes. In-situ monitoring data is a dynamic result of the combined actions of a plurality of factors. The change in ambient temperature affects the dielectric parameters of the insulating material, and its effect overlaps with the characteristic change caused by moisture in the time domain. Random disturbance such as electromagnetic interference and mechanical stress can be continuously mixed into the monitoring signal to form noise and mutation. The prior art can not effectively separate three components with different properties, namely a damping characteristic, a temperature effect and instantaneous disturbance, from a continuous data stream, so that the system is insensitive to early weak damping signals and is easy to generate false alarm due to interference. The existing evaluation model usually processes humidity related parameters in an isolated manner or adopts fixed temperature and humidity compensation coefficients, lacks quantitative description of dynamic correlation between humidity and temperature signals, and does not consider the influence of random disturbance on the correlation. The evaluation result is difficult to adapt to complex running environment changes, the real risk level corresponding to the same monitoring value under different working conditions cannot be accurately reflected, and the early warning accuracy and reliability are limited. A method capable of performing dynamic feature decoupling on a monitored data stream and establishing a self-adaptive probability evaluation model is needed to realize accurate identification of the moisture state of the sleeve. Disclosure of Invention The invention aims to provide a method and a system for monitoring the moisture of a capacitive transformer bushing, which are used for solving the problems in the background technology. In order to achieve the above object, the present invention provides a method for monitoring moisture in a bushing of a capacitive transformer, the method comprising: acquiring an initial monitoring data stream from a monitoring node deployed in a transformer bushing; executing a feature extraction strategy based on time evolution analysis on the initial monitoring data stream to obtain a preliminary humidity response sequence, a preliminary temperature response sequence and a potential disturbance identification sequence; performing relevance verification of the preliminary humidity response sequence and the preliminary temperature response sequence, and calculating coupling influence degree of the potential disturbance identification sequence on the preliminary humidity response sequence and the preliminary temperature response sequence; constructing a dynamic probability response matrix according to the correlation verification result and the coupling influence degree, and deducing a damping evaluation index based on the dynamic probability response matrix; And carrying out multi-level signal correction on the initial monitoring data flow according to the damping evaluation index so as to analyze a core humidity component and a core temperature component reflecting the damping process. Preferably, the performing a feature extraction strategy based on time evolution analysis on the initial monitoring data stream includes the following steps: constructing a sliding evaluation window for the initial monitoring data stream; identifying and dividing a data segment with continuous stationary attribute and a data segment with transient non-stationary attribute in the initial monitoring data stream in the sliding evaluation window; For the data segment with continuous and stable attribute, adopting a trend stability detection method to separate a baseline humidity component and a baseline temperature component, wherein the baseline humidity component and the baseline temperature component form a preliminary humidity response sequence and partial basic data of the preliminary temperature response sequence; And starting an abnor