CN-121978576-A - Distribution box electric leakage monitoring and early warning system

CN121978576ACN 121978576 ACN121978576 ACN 121978576ACN-121978576-A

Abstract

The invention particularly relates to a distribution box electric leakage monitoring and early warning system, which relates to the technical field of power equipment monitoring and comprises a multi-mode sensing array module, a heterogeneous signal processing and fusing module and an intelligent diagnosis and three-dimensional positioning engine module. According to the multi-mode fusion sensing technology, early insulation degradation characteristics such as electric field distortion and radio frequency pulse are captured, a prediction model with a attention mechanism is combined, an insulation degradation risk value of 24 hours in the future is output, and the temperature and humidity compensation and anti-interference filtering design is matched, so that false alarm and missing report caused by power grid harmonic waves and electromagnetic radiation are effectively avoided, the span from post-working alarm to pre-prediction is realized, and sufficient treatment time is reserved for operation and maintenance.

Inventors

ZHAO XUJIE
ZHAO PENGJIE
WANG RUJIE
SHE LIJUN

Assignees

浙江章开电器有限公司

Dates

Publication Date: 20260505
Application Date: 20260122

Claims (9)

1. The utility model provides a block terminal electric leakage monitoring early warning system which characterized in that includes: the multi-mode sensing array module is configured to adopt a distributed array and integrated probe mixed deployment and synchronously collect electric field distortion, radio frequency/ultrahigh frequency pulse and current/temperature/vibration signals; the heterogeneous signal processing and fusion module is configured to realize deep fusion of heterogeneous signals through a special conditioning circuit noise reduction, multidimensional feature extraction and space-time alignment technology, and generate a joint feature vector; The intelligent diagnosis and three-dimensional positioning engine module is configured to realize early prediction of insulation degradation, fault type classification and three-dimensional positioning based on a Bi-LSTM prediction model, a CNN and SVM classifier and an electric field-radio frequency fusion positioning algorithm.
2. The electrical box leakage monitoring and early warning system according to claim 1, wherein the multi-mode sensing array module specifically comprises: a space electric field distortion perception sub-module: the MEMS electric field sensor is adopted to deploy a preset number of sensors on the inner wall of the distribution box according to grid layout After the system is electrified for the first time, continuously acquiring electric field data in a healthy state, storing amplitude and phase information of each sensor, and generating a reference three-dimensional electric field distribution map; During normal operation, the sensor array collects data in real time, and each frame of data is compared with the corresponding position of the reference map to calculate the electric field distortion coefficient.
3. The electrical box leakage monitoring and early warning system according to claim 2, further comprising a radio frequency/ultra-high frequency detection and injection sub-module: The microstrip patch antennas are adopted to arrange 5 antennas according to the four corners and the center layout, and the antennas are respectively positioned at the four corners and the center position of the top of the distribution box; Passive mode: The antenna array receives the space high-frequency electromagnetic pulse in real time, filters clutter after being amplified by the low-noise amplifier, and then converts the clutter into digital signals; carrying out spectrum analysis on the digital signal, extracting the energy duty ratio, pulse repetition rate and pulse amplitude distribution of the partial discharge characteristic frequency band, and preliminarily judging a discharge area by firstly capturing antenna identification and signal arrival time difference; Active mode: the signal generator generates sine wave scanning signals, and the sine wave scanning signals are amplified by the power amplifier and then injected into the bus through the coupler; When the insulation resistance is reduced, the impedance to ground is reduced, the reflection coefficient is increased, the transmission attenuation is reduced, and the insulation resistance variation is quantified by comparing the response curves under the healthy state; The passive mode output comprises a high-frequency pulse time domain waveform, a spectrogram, a PRPD (pulse-with-pulse-duration) spectrogram, a characteristic frequency band energy duty ratio and a preliminary positioning area; the active mode outputs are incident/reflected/transmitted signal waveforms, impedance-frequency response curves, reflection coefficient-frequency curves, insulation resistance estimates.
4. The electrical box leakage monitoring and early warning system according to claim 1, further comprising a multiparameter fusion probe sub-module: The core sensor integration comprises a miniature rogowski coil, a high-precision temperature sensor, a micro-vibration sensor and UHF antenna contacts, wherein the miniature rogowski coil, the high-precision temperature sensor, the micro-vibration sensor and the UHF antenna contacts are nested on a bus and a cable through a buckle structure; Controlling each sensor to synchronously acquire data, wherein a sampling trigger signal is provided by a system unified clock; The method comprises the steps of carrying out primary processing on collected original data, including harmonic analysis of current signals, moving average filtering of temperature signals, peak detection of vibration signals, and outputting synchronous data frames and abnormal triggering data.
5. The distribution box leakage monitoring and early warning system according to claim 1, wherein the heterogeneous signal processing and fusing module specifically comprises: signal conditioning and noise reduction dedicated circuitry: For the signal characteristics of different types of sensors, a modularized conditioning circuit is designed, and each module is independently packaged: Electric field sensor signal conditioning circuit: an input stage, namely a high-impedance operational amplifier is adopted; an amplifying stage adopts instrumentation amplifier; A filtering stage, namely a second-order active low-pass filter is adopted; an output stage, which adopts a voltage follower; Radio frequency signal conditioning circuit: a passive signal channel, including a broadband low noise amplifier, a programmable band-pass filter and a variable gain amplifier; The active injection signal channel comprises a signal generator, a power amplifier, a directional coupler and a reflected signal receiving amplifier; Fusion probe signal conditioning circuit: The current signal channel is used for converting the output signal of the Rogowski coil into a voltage signal through an integrator and conditioning the voltage signal through a low-pass filter and an amplifier; a temperature signal channel is driven by a constant current source and amplifies a voltage signal through an instrument amplifier; and the vibration signal channel is used for extracting a vibration peak value signal from the preamplifier, the band-pass filter and the peak value holding circuit.
6. The electrical box leakage monitoring and early warning system of claim 5, further comprising feature extraction: electric field data feature extraction: Calculating the spatial gradient, the equipotential surface torsion degree and the distortion area of the electric field distribution; extracting the mean value, standard deviation, maximum value, change rate and peak factor of the electric field distortion coefficient; And (3) extracting radio frequency data characteristics: Calculating the energy duty ratio of a characteristic frequency band, the mass center of the frequency spectrum and the flatness of the frequency spectrum; extracting pulse repetition rate, pulse amplitude distribution entropy, pulse rising time and pulse width; fusion probe data feature extraction: Calculating effective value, peak value, form factor, harmonic distortion rate, amplitude and phase of each subharmonic; Extracting temperature average value, maximum value, temperature change rate and temperature fluctuation amplitude; extracting characteristic frequency, harmonic content and vibration energy of a vibration signal through FFT conversion; and calculating a correlation coefficient of the current and the temperature and a time difference between the vibration peak value and the current peak value.
7. The distribution box leakage monitoring and early warning system according to claim 1, wherein the intelligent diagnosis and three-dimensional positioning engine module specifically comprises: Early prediction model of insulation degradation: a bidirectional LSTM network is adopted, and comprises 3 hidden layers, wherein each hidden layer has 128 neurons, the input dimension is 10 dimensions, and the output dimension is 1 dimension; adding key features with large influence on a prediction result by automatic focusing of an attention mechanism after an LSTM layer, and mapping a risk value to a 0-1 interval; the data set construction, namely collecting actual measurement data of different insulation degradation stages, wherein each group of samples comprises 24 hours of time sequence data, expanding the sample size, and dividing the sample size into a training set, a verification set and a test set according to the proportion of 7:2:1; The training parameters adopt an Adam optimizer; inputting the latest 2-hour time sequence characteristic every 5 minutes, and outputting an insulation degradation risk value curve of 24 hours in the future by a model; analyzing the slope of the risk value curve through a linear fitting algorithm; Fault type accurate classifier: The feature input layer is used for inputting 15-dimensional fusion feature vectors; The characteristic enhancement layer is used for extracting local characteristics by adopting a 1D convolution layer, and the activation function is ReLU; The feature aggregation layer adopts a maximum pooling layer to reduce feature dimension, and then carries out feature aggregation through the full connection layer; The classification output layer is used for outputting probability distribution of 6 types of faults by adopting an SVM classifier; And constructing a data set, namely acquiring actual measurement data of 6 types of faults, wherein each type of faults acquires a preset number of groups of samples, and each group of samples comprises 15-dimensional fusion characteristics.
8. The electrical box leakage monitoring and early warning system according to claim 1, further comprising: Solving and positioning based on inverse problem of electric field array: establishing a finite element model of three-dimensional electric field propagation in the distribution box, and considering conductor shape, dielectric constant of insulating materials and air medium; Solving an inverse problem, namely taking an electric field distortion vector matrix as input, and solving the inverse problem through a regularized least square algorithm to obtain equivalent charge distribution of the insulation defect; The barycentric coordinates of the equivalent charge distribution are the preliminary positioning coordinates of the fault points, and are based on the time difference positioning of the radio frequency antenna array, the time difference of the high frequency pulse reaching 5 antennas is calculated through a cross correlation algorithm; based on the known coordinates of the antenna array, an arrival time difference positioning equation set is established, the equation set is solved to obtain the positioning coordinates of the fault points, and the final positioning coordinates are calculated by adopting weighted average fusion.
9. The electrical box leakage monitoring and early warning system according to claim 1, further comprising: The grading early warning and self-calibration output module is configured to adopt a four-level dynamic early warning strategy to issue fault information in multiple channels, and combines a periodic self-calibration, sensor health monitoring and degradation operation mechanism.

Description

Distribution box electric leakage monitoring and early warning system Technical Field The invention relates to the technical field of power equipment monitoring, in particular to a distribution box leakage monitoring and early warning system. Background The distribution box is used as core equipment for distributing and controlling electric energy in an electric power system, and the insulation performance of the distribution box directly determines the safety and stability of electric power supply. When insulation degradation (such as damp, carbonization and aging) occurs in the distribution box, leakage faults are easy to occur, and if the monitoring and early warning are not performed in time, serious accidents such as equipment burning, fire disaster and even personnel electric shock can be caused. The existing distribution box leakage monitoring technology mainly relies on a single current transformer to collect leakage current signals, and has the following remarkable defects: The early warning hysteresis is strong, the warning can be triggered only when the insulation degradation forms a stable leakage channel and the leakage current reaches a threshold value, and the early fine characteristics of the insulation degradation can not be captured; The positioning accuracy is poor, only the leakage of the loop can be judged, the position of the fault point in the box cannot be accurately positioned, and great inconvenience is brought to operation and maintenance; The diagnosis depth is insufficient, the physical types of leakage faults (such as corona discharge, air gap discharge and the like) cannot be distinguished, the targeted operation and maintenance are difficult to support, and fourth, the anti-interference capability is weak, and the fault diagnosis device is easily influenced by power grid harmonic waves and environmental electromagnetic radiation, so that false alarm or missing alarm is caused. Therefore, developing a distribution box monitoring system capable of realizing early warning, accurate positioning and deep diagnosis of electric leakage faults becomes a technical problem to be solved in the field of power operation and maintenance. Disclosure of Invention The invention aims to solve the problems and provides a distribution box leakage monitoring and early warning system. In order to achieve the above purpose, the present invention adopts the following technical scheme: an electrical distribution box leakage monitoring and early warning system, comprising: the multi-mode sensing array module is configured to adopt a distributed array and integrated probe mixed deployment and synchronously collect electric field distortion, radio frequency/ultrahigh frequency pulse and current/temperature/vibration signals; the heterogeneous signal processing and fusion module is configured to realize deep fusion of heterogeneous signals through a special conditioning circuit noise reduction, multidimensional feature extraction and space-time alignment technology, and generate a joint feature vector; The intelligent diagnosis and three-dimensional positioning engine module is configured to realize early prediction of insulation degradation, fault type classification and three-dimensional positioning based on a Bi-LSTM prediction model, a CNN and SVM classifier and an electric field-radio frequency fusion positioning algorithm. Preferably, the multi-mode sensing array module specifically includes: a space electric field distortion perception sub-module: deploying a preset number of sensors on the inner wall of the distribution box according to grid layout by adopting MEMS electric field sensors; After the system is electrified for the first time, continuously acquiring electric field data in a healthy state, storing amplitude and phase information of each sensor, and generating a reference three-dimensional electric field distribution map; During normal operation, the sensor array collects data in real time, and each frame of data is compared with the corresponding position of the reference map to calculate the electric field distortion coefficient. Preferably, the method further comprises a radio frequency/ultrahigh frequency detection and injection sub-module: The microstrip patch antennas are adopted to arrange 5 antennas according to the four corners and the center layout, and the antennas are respectively positioned at the four corners and the center position of the top of the distribution box; Passive mode: The antenna array receives the space high-frequency electromagnetic pulse in real time, filters clutter after being amplified by the low-noise amplifier, and then converts the clutter into digital signals; carrying out spectrum analysis on the digital signal, extracting the energy duty ratio, pulse repetition rate and pulse amplitude distribution of the partial discharge characteristic frequency band, and preliminarily judging a discharge area by firstly capturing antenna identification and sig