CN-122026948-A - Automatic fault section identification and remote isolation method for low-voltage distribution line

CN122026948ACN 122026948 ACN122026948 ACN 122026948ACN-122026948-A

Abstract

The invention relates to the technical field of circuit fault processing, and discloses an automatic fault section identification and remote isolation method of a low-voltage distribution line, wherein a reference fingerprint library based on channel state information is established through a central coordination node; the method comprises the steps of monitoring real-time channel frequency response, executing threshold comparison to judge suspected fault points, triggering an active impedance control unit of a terminal proxy node to execute controlled perturbation, calculating theoretical transfer function variation by utilizing a transfer line network model, carrying out fitting analysis on the theoretical transfer function variation and an actual measured value, judging physical faults through normalized residual error coefficients, stripping load interference, calculating residual error drift quantity between adjacent nodes based on a topology mapping table to execute differential positioning to lock fault physical intervals, and sending trip messages to drive a circuit breaker to execute physical isolation at a power frequency zero crossing point. According to the scheme, the on-line fault identification and isolation are realized through active impedance perturbation and distributed residual analysis, and the robustness of fault judgment under a complex load environment is improved.

Inventors

REN WENJI
QIAO JIAQI
ZHANG JIAN
HU PENGFEI
ZHOU XIAOCHUANG
LIU MINGDONG

Assignees

河南北筑沥青发泡科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260224

Claims (10)

1. The automatic fault section identification and remote isolation method for the low-voltage distribution line is characterized by comprising the following steps of: s100, a central coordination node instructs a whole network terminal proxy node to enter a reference test state, and the terminal proxy node extracts physical layer channel state information and establishes a pre-stored channel fingerprint library by the central coordination node; s200, a central coordination node monitors a power line carrier channel in real time, and a terminal agent node extracts a real-time channel state information matrix and sends the real-time channel state information matrix to the central coordination node; S300, the central cooperative control point compares the real-time frequency response vector with a reference frequency response vector in a pre-stored channel fingerprint library in a numerical mode, and when the Euclidean distance between the real-time frequency response vector and the reference frequency response vector exceeds a preset abnormal deviation threshold value, a suspected fault point is judged to exist; S400, an active impedance control unit of a terminal agent node of a central coordination node triggering target area executes impedance perturbation action, the active impedance control unit changes input impedance of the terminal agent node by switching a precise impedance element matrix, and the central coordination node synchronously captures channel state information secondary fluctuation characteristics caused by active impedance switching; S500, the central coordination node performs logic verification by utilizing the secondary fluctuation characteristic of the channel state information, performs fitting analysis on the actually measured transfer function variation and the theoretical transfer function variation calculated based on a transfer line equation, and determines that physical faults occur on the power line when fitting residual errors are smaller than a judgment boundary; And S600, the central coordination node performs distributed cooperative positioning and section isolation, locks a fault physical interval by calculating residual drift quantity between adjacent terminal proxy nodes, and transmits a trip isolation message to drive a breaker to perform physical link isolation.
2. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 1, wherein the process of extracting the physical layer channel state information in step S100 comprises: The terminal agent node locates pilot frequency sub-carriers in the orthogonal frequency division multiplexing symbol sequence, and calculates the ratio of the received frequency domain complex signals to preset pilot frequency complex symbols by utilizing a least square estimation method to obtain the frequency response value of the pilot frequency sub-carrier position; And the terminal agent node complements the frequency response estimated value of the data subcarrier position through a linear interpolation algorithm, and adopts a sliding window smoothing algorithm to carry out secondary filtering on the continuous channel frequency response estimated value so as to generate the physical layer channel state information covering the full frequency band.
3. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 1, wherein the process of executing the impedance perturbation action by the active impedance control unit in step S400 includes: the power line carrier chip of the terminal agent node outputs binary coding signals to the isolation driving circuit through a general input/output pin; The isolation driving circuit drives the switching circuit to control the two-way metal oxide semiconductor field effect transistor in the array to be conducted, and the controlled impedance branch matrix is connected in parallel to the secondary side of the power line coupling loop; the active impedance control unit finishes level inversion at the starting moment of the cyclic prefix of the orthogonal frequency division multiplexing symbol, so that the equivalent input impedance of the active impedance control unit generates step change.
4. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 1, wherein the calculating the theoretical transfer function variation based on the transfer line equation in step S500 comprises: The central coordination node establishes a two-port network ABCD matrix model, and matrix parameters in the two-port network ABCD matrix model are determined by resistance per unit length, inductance per unit length, conductance per unit length and capacitance per unit length of the low-voltage distribution line; the central coordination node calculates the theoretical transfer function variation caused by the preset impedance increment generated by the active impedance control unit by using a reference frequency response function as an initial value.
5. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 4, wherein said performing said fitting analysis in step S500 comprises: The central coordination node calculates a residual vector between the actually measured transfer function variable quantity and the theoretical transfer function variable quantity; The central coordination node calculates a normalized residual coefficient in the full-band range according to the residual vector, and compares the normalized residual coefficient with a fault judgment threshold; and the fault judgment threshold value is dynamically adjusted according to the signal-to-noise ratio of the current power line carrier channel.
6. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 5, wherein said logic verification in step S500 further comprises the substep of eliminating abrupt load disturbance: The central coordination node calculates a cross-correlation function between an impedance switching instruction issuing moment mark and a starting moment mark of the channel state information fluctuation, and if the peak displacement of the cross-correlation function exceeds a preset time window threshold, the channel state information fluctuation is judged to be caused by random load fluctuation; And the central coordination node extracts a phase deviation angle of the residual vector, and if the phase deviation angle does not accord with the complex impedance characteristic phase preset by the active impedance control unit, the random load fluctuation is filtered.
7. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 1, wherein said step S600 of calculating said residual drift between adjacent said terminal agent nodes comprises: The central coordination node determines a logic connection level of each terminal agent node in the low-voltage distribution line according to an internally stored topology mapping table; The central coordination node calculates absolute difference values of the normalized residual coefficients of the adjacent upstream terminal agent nodes and the normalized residual coefficients of the downstream terminal agent nodes in the topology mapping table, and the residual drift amount is obtained; When the residual drift amount exceeds a segment failure threshold, the central coordination node determines that a failure occurred in the failed physical section between the upstream and downstream end proxy nodes.
8. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 7, wherein said step S600 of driving said circuit breaker to perform physical link isolation comprises: The central coordination node transmits the tripping isolation message containing a message authentication code field and a unidirectional counter field to a target terminal proxy node; And after the target terminal agent node verifies that the tripping isolation message is legal, acquiring a power frequency voltage zero crossing point by using an internal zero crossing detection circuit, and driving the circuit breaker to execute breaking action at the moment of the power frequency voltage zero crossing point.
9. The method for automatically identifying and remotely isolating a fault section of a low voltage power distribution line according to claim 1, wherein said step S600 further comprises the sub-steps of verifying and locking: After the terminal agent node confirms that the breaker is disconnected by reading an auxiliary contact state register, feeding back an isolation completion signal to the central coordination node; The central coordination node marks the fault physical interval as an offline state in the topology mapping table and executes the logic locking; And if the residual drift amount of three continuous periods does not stably exceed the section fault threshold, the central coordination node prohibits issuing the tripping isolation message.
10. A low voltage distribution line fault active identification system for implementing the method of any one of claims 1 to 9, the low voltage distribution line fault active identification system comprising: The central coordination node is deployed on a low-voltage side bus of the distribution transformer and is used for maintaining a topology mapping table, calculating a normalized residual error coefficient and generating a tripping isolation message; The terminal agent nodes are deployed at the distribution branch line access points and interact with the central coordination node through a power line carrier network; the active impedance control unit is integrated inside the terminal agent node and comprises a switching circuit control array formed by bidirectional metal oxide semiconductor field effect transistors and a precise impedance element matrix, and is used for generating controlled impedance perturbation signals; And the circuit breaker is controlled by the terminal agent node and is used for receiving the tripping isolation message and executing physical link isolation operation.

Description

Automatic fault section identification and remote isolation method for low-voltage distribution line Technical Field The invention relates to the technical field of circuit fault processing, in particular to an automatic fault section identification and remote isolation method for a low-voltage distribution line. Background The fault monitoring and diagnosis of the low-voltage distribution network is a key link for guaranteeing the reliability of power supply. The power line carrier communication technology uses the existing distribution cable as a signal transmission medium, and can acquire channel state information reflecting the evolution of a line topological structure and physical parameters by collecting the transmission characteristics of a high-frequency carrier signal in a physical link in real time. Such information is typically obtained by demodulation of the physical layer pilot signal of an orthogonal frequency division multiplexing system, providing a data basis for identifying changes in physical properties such as line grounding, line breakage, insulation degradation, and the like. The prior art mainly adopts a passive detection mode in low-voltage distribution line monitoring. And periodically reading channel quality parameters or electrical characteristic data sent by each access terminal by distributing monitoring nodes at the outlet side of the distribution transformer. When the topology change is caused by the physical fault of the line, the amplitude-frequency characteristic and the phase-frequency characteristic of the carrier signal can generate random fluctuation, and the monitoring system identifies the abnormality of the running state of the line by carrying out threshold judgment or statistical analysis on the passively acquired channel fluctuation data. In addition, the correlation verification is carried out on the acquired real-time channel response data by combining a part of schemes with a method of comparing the historical fingerprint database, so as to assist in judging the stability of the line state. The existing passive monitoring scheme faces false alarm recognition dilemma caused by feature confusion in practical application. Because the low-voltage distribution network access load has high time variability and randomness, the high-power electric equipment is frequently started or stopped, and channel response fluctuation caused by nonlinear load access shows high similarity with channel distortion generated by early-stage line physical damage on amplitude characteristics and phase characteristics. Because the monitoring system lacks a physical active induction and verification means for channel fluctuation, a deterministic causal mapping relation cannot be established, so that the system is difficult to accurately filter characteristic coupling between random load interference and line real faults at a physical level, and the accuracy of fault judgment is limited. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a fault section automatic identification and remote isolation method of a low-voltage distribution line, which solves the problems mentioned in the background art. The invention provides a method for automatically identifying and remotely isolating a fault section of a low-voltage distribution line, which is cooperatively executed by a central coordination node and a plurality of terminal agent nodes deployed at the tail ends of distribution branch lines and senses the evolution of the impedance characteristics of a power grid by utilizing the channel state information of a physical layer of a power line carrier communication system. The method first performs reference fingerprint establishment. The central coordination node instructs the whole network terminal proxy node to extract the physical layer channel state information under the normal operation condition. The terminal agent node calculates the frequency response function through the pilot frequency sub-carrier wave of the orthogonal frequency division multiplexing technologyThe calculation mode is as follows: In the formula (i) the formula (ii), For the index of the sub-carrier,For the frequency domain complex signal acquired by the receiving end,Pilot complex symbols preset for the transmitting end. And the central coordination node establishes a pre-stored channel fingerprint library according to the acquired frequency response function. And then enters a real-time monitoring stage. And the central coordination node periodically acquires a real-time frequency response vector uploaded by the terminal agent node, and calculates the Euclidean distance between the real-time frequency response vector and corresponding reference data in a pre-stored channel fingerprint library. When the Euclidean distance exceeds a preset abnormal deviation threshold value, the system judges that a suspected fault point exists in the distribution line. The innovation part of the pres