Search

CN-121978467-A - Holographic diagnosis fault positioning system for low-voltage distribution line

CN121978467ACN 121978467 ACN121978467 ACN 121978467ACN-121978467-A

Abstract

The invention discloses a holographic diagnosis fault positioning system for a low-voltage distribution line, which relates to the field of holographic diagnosis and comprises the steps of acquiring line electrical, environment, physical state and equipment operation parameters in real time through a data acquisition module, removing noise and supplementing data, carrying out precision correction by a calibration module based on a standard model library, integrating multi-source standardized data through a three-level progressive strategy by a holographic data fusion module, identifying fault characteristics from the fused data through a fault characteristic extraction module in combination with deep learning, realizing multidimensional diagnosis through a holographic diagnosis analysis module by integrating three large models of mechanism analysis, machine learning and rule reasoning, carrying out fault point calculation by utilizing space-time characteristics, and ensuring reliability of a diagnosis result through a four-level closed-loop verification mechanism. The invention has the advantages that by means of full-link multidimensional holographic data support, comprehensive diagnosis and reliable positioning of faults of the low-voltage distribution line are realized through multi-model collaborative diagnosis and multi-algorithm fusion positioning and matching with four-level closed-loop verification.

Inventors

  • ZHAO BINGBING

Assignees

  • 哈尔滨华创弦柯电气有限公司

Dates

Publication Date
20260505
Application Date
20260327

Claims (9)

  1. 1. A low voltage distribution line holographic diagnostic fault location system, comprising: The data acquisition module is used for acquiring the full-link and multidimensional operation data of the low-voltage distribution line in real time, constructing holographic data for fault diagnosis, and acquiring the electrical parameters, the environmental parameters, the physical state parameters and the operation state parameters of equipment of the line in a coverage range; the data preprocessing module is in signal connection with the data acquisition module and is used for carrying out noise rejection, redundant filtration and integrity completion on the acquired original data; The data calibration module is internally provided with a multi-dimensional standard calibration model library, corresponding calibration parameters and correction algorithms are preset for different types of acquired data, and the preprocessed clean data are subjected to precision calibration and deviation correction; the holographic data fusion module is used for carrying out cross-dimension and full-element holographic fusion processing on the calibrated standardized data by adopting a three-level progressive holographic fusion strategy, and constructing a holographic data set for comprehensively describing the running state of the line; The fault feature extraction module is used for constructing a multi-type fault feature template library and extracting feature data related to various faults from the holographic fusion data by adopting a mixed feature extraction method combining deep learning and traditional signal processing; The holographic diagnosis analysis module integrates an analysis model based on a fault mechanism, a machine learning diagnosis model based on deep learning and a rule reasoning model based on expert experience, and performs full-dimensional and multi-view holographic diagnosis of the fault of the low-voltage distribution line through cooperative linkage of the three models; the accurate positioning calculation module adopts a positioning strategy fused by multiple algorithms, completes positioning calculation by combining time sequence characteristics and space characteristics of holographic data, and realizes accurate positioning of fault points; And the diagnosis result verification module adopts a four-level closed loop verification mechanism to verify the diagnosis result and the fault positioning result in a multi-dimensional and full-link manner.
  2. 2. The low voltage distribution line holographic diagnostic fault location system of claim 1, wherein said data acquisition module comprises: The multi-type sensing unit is used for collecting line electrical parameters, environment parameters, line physical state parameters and equipment operation state parameters, and specifically comprises three-phase voltage, three-phase current, zero-sequence current, line active power, reactive power and power factor electrical parameters, temperature, humidity, wind speed, precipitation and ice coating thickness environment parameters, line joint temperature, insulator pollution degree, conductor sag line physical state parameters and operation state parameters of a distribution transformer, a circuit breaker and a fuse device; the hybrid communication unit adopts LoRa ad hoc network transmission at a short distance and accesses a distribution network backbone network at a long distance through an Ethernet; and the clock synchronization unit is used for integrating the Beidou time service module and dynamically calibrating the time stamp of each acquisition terminal through the NTP protocol.
  3. 3. The low voltage distribution line holographic diagnostic fault location system of claim 1, wherein said data preprocessing module specifically comprises: The noise removing unit is used for identifying abnormal noise data through a multi-threshold screening mechanism, adaptively adjusting a filtering window according to the fluctuation intensity of the data and removing noise caused by electromagnetic interference of a distribution line; the data normalization unit is used for mapping the data to a [0,1] interval by distinguishing dimension attributes of different types of data of electric, environment and equipment states and recording dimension conversion mapping relations by adopting a hierarchical scaling strategy; The time sequence interpolation complement unit is used for complementing the missing data by adopting a segmentation interpolation strategy based on the historical change trend of the data and the synchronous similar working condition data; And the data quality checking unit is used for constructing an integrity, consistency and accuracy three-dimensional checking index system, performing full-quantity checking on the preprocessed data, and triggering a backtracking reprocessing flow by unqualified data.
  4. 4. The low voltage distribution line holographic diagnostic fault location system of claim 1, wherein said data calibration module comprises: the standard calibration model library unit is internally provided with three main calibration models of electricity, environment and equipment state, preset calibration reference parameters corresponding to different acquisition terminal models and line working conditions, and store historical calibration data and deviation correction data; The scene matching and model selecting unit is used for extracting acquisition scene information and acquisition terminal identification information of the preprocessed data and matching a corresponding standard calibration model; The deviation correction unit calculates the deviation between the measured data and the standard model reference value in real time, dynamically adjusts the correction coefficient by combining the operation time of the acquisition terminal and the change of the environmental temperature and humidity, and completes deviation compensation; and the calibration result checking unit is used for constructing double check indexes of precision and consistency, performing full-quantity check on the calibrated data, triggering a backtracking recalibration process by unqualified data, and attaching a calibration precision rating label to the qualified data.
  5. 5. The low-voltage distribution line holographic diagnostic fault location system of claim 1, wherein said holographic data fusion module comprises: The data layer fusion unit is used for carrying out time axis alignment and redundancy elimination on the standardized data of different acquisition nodes of the same type and generating a continuous data sequence covering the whole link of the line through a sliding window aggregation algorithm; The feature layer fusion unit is used for extracting time domain, frequency domain and time domain features of each dimension data, establishing cross-dimension feature association through feature correlation analysis and generating a unified feature set; the decision layer fusion unit is used for carrying out weight distribution and conflict resolution on the primary state judgment result of each dimension data, and outputting a comprehensive state evaluation result after integrating the complementary information; And the fusion process control unit is used for coordinating the cooperative work of the three-level fusion units, recording fusion process parameters and triggering a backtracking re-fusion mechanism for fusion abnormal data.
  6. 6. The low voltage distribution line holographic diagnostic fault location system of claim 1, wherein said fault signature extraction module comprises: The fault feature template library unit is internally provided with a short circuit, grounding and overload fault feature subset and is used for storing typical time domain mutation threshold values, frequency domain distortion intervals and feature change time sequence rules corresponding to each fault and associating fault causes with feature mapping relations; The multi-domain feature mining unit integrates a time domain, a frequency domain and a deep abstract feature extraction submodule, the time domain submodule captures a data mutation peak value and duration, the frequency domain submodule identifies fault frequency distortion features, and the deep submodule mines non-explicit association features; the feature matching and screening unit is used for precisely matching the multi-domain mining features with the template library by adopting a feature similarity comparison algorithm to generate a targeted fault feature vector; And the characteristic quality control unit is used for carrying out integrity and consistency check on the generated characteristic vector, and triggering a backtracking and re-mining flow by unqualified characteristics.
  7. 7. The low-voltage distribution line holographic diagnostic fault location system of claim 1, wherein said holographic diagnostic analysis module specifically comprises: the multi-model fusion diagnosis engine unit is used for integrating three sub-modules, namely an analysis model, a machine learning diagnosis model and a rule reasoning model, wherein the analysis model is used for constructing a logic judgment link based on a fault mechanism, locking a fault type range, the machine learning model is used for carrying a deep training network, mining holographic data and hiding associated information, and the rule reasoning model is used for integrating an expert experience library in the field to form a complementary diagnosis link; The model collaborative scheduling unit adopts a dynamic weight distribution mechanism, adjusts the output weight of each model according to the complexity of the fault scene, integrates the difference conclusion through a conflict resolution algorithm, and outputs a unified diagnosis result; And the learning and updating unit is used for absorbing newly-increased fault data and diagnosis experience in real time and dynamically optimizing model parameters and weight configuration.
  8. 8. The low-voltage distribution line holographic diagnostic fault location system of claim 1, wherein the precise location calculation module specifically comprises: the primary positioning calculation unit is used for constructing a fault impedance analysis link based on the electrical parameters and the line inherent parameters in the holographic fusion data and obtaining the initial distance between a fault point and an acquisition node; The positioning result correction unit is used for calculating traveling wave propagation speed and time difference based on propagation time and reflection characteristics of fault traveling waves, performing deviation compensation on the preliminary positioning result and obtaining corrected accurate distance data; the space topology mapping unit is used for converting the corrected fault distance into geographic coordinates, associating corresponding line sections and equipment identification information, generating distance-coordinate-equipment three-dimensional association data and accurately mapping fault point space; and the complex scene adapting unit dynamically matches the positioning algorithm parameters of the corresponding scene aiming at line branches and crossing complex scenes.
  9. 9. The low-voltage distribution line holographic diagnostic fault location system of claim 1, wherein said diagnostic result verification module specifically comprises: The data consistency verification unit is used for extracting a change curve of the holographic fusion data before and after the fault, comparing the matching degree of the diagnosis conclusion with the mutation characteristics and the trend abnormality of the data, screening unqualified diagnosis results and marking deviation types; and the model cross verification unit re-analyzes the same fault data through analysis, machine learning and rule reasoning models of the holographic diagnosis core, calculates consistency coefficients of diagnosis conclusions of all models and verifies diagnosis stability. The space association verification unit is used for carrying out field association verification on the fault point positioning result by combining the line real-time monitoring image, the unmanned aerial vehicle inspection data and the equipment running state feedback information; and the historical case verification unit is used for comparing the characteristic vector, diagnosis and positioning information of the current fault with the historical fault case library, analyzing the relevance of fault types, causes, positions and influence factors, and verifying the rationality of diagnosis and positioning results.

Description

Holographic diagnosis fault positioning system for low-voltage distribution line Technical Field The invention relates to the field of holographic diagnosis, in particular to a holographic diagnosis fault positioning system for a low-voltage distribution line. Background Along with the rapid increase of novel loads such as distributed energy sources and electric automobiles, the traditional power distribution network has the problems of high fault rate, low positioning efficiency, enlarged power failure influence range and the like. According to statistics, the faults of the distribution network account for more than 80% of the accidents of the power system, and the traditional methods such as manual inspection, sectional trial power transmission and the like need 2-4 hours to locate fault points on average, so that the power supply reliability is seriously affected. The core disadvantages of the low-voltage distribution line fault diagnosis and positioning system in the current market are concentrated on four dimensions of data perception, fusion analysis, positioning accuracy and verification mechanism. The data acquisition is limited to single electrical parameters, the environment, the physical state of the line and other multidimensional data are lacked, the fusion technology is crude, cross-dimension space-time alignment and depth association are difficult to realize, and a comprehensive line operation holographic view cannot be constructed. The diagnosis model is dependent on a single algorithm or a traditional signal processing method, lacks a multi-model cooperative linkage mechanism, has low recognition rate on complex faults and high-resistance faults, and is easy to be interfered by complex scenes such as distributed power supply access. The positioning algorithm has poor adaptability, has larger error under complex topologies such as branch lines, mixed lines and the like, and is difficult to realize accurate positioning. Meanwhile, a perfect full-link verification mechanism is generally lacking, reliability of diagnosis and positioning results is not guaranteed, manual assistance and research are relied on, fault processing efficiency is low, and accurate operation and maintenance requirements of a complex power distribution network are difficult to meet. Disclosure of Invention In order to perfect the existing system, the method provides a low-voltage distribution line holographic diagnosis fault positioning system, and the method relies on full-link multidimensional holographic data support, and realizes comprehensive accurate diagnosis and reliable positioning of low-voltage distribution line faults through multi-model collaborative diagnosis and multi-algorithm fusion positioning and matching with four-level closed loop verification. In order to achieve the above purpose, the invention adopts the following technical scheme: A low voltage distribution line holographic diagnostic fault location system comprising: The data acquisition module is used for acquiring the full-link and multidimensional operation data of the low-voltage distribution line in real time, constructing holographic data for fault diagnosis, and acquiring the electrical parameters, the environmental parameters, the physical state parameters and the operation state parameters of equipment of the line in a coverage range; the data preprocessing module is in signal connection with the data acquisition module and is used for carrying out noise rejection, redundant filtration and integrity completion on the acquired original data; The data calibration module is internally provided with a multi-dimensional standard calibration model library, corresponding calibration parameters and correction algorithms are preset for different types of acquired data, and the preprocessed clean data are subjected to precision calibration and deviation correction; the holographic data fusion module is used for carrying out cross-dimension and full-element holographic fusion processing on the calibrated standardized data by adopting a three-level progressive holographic fusion strategy, and constructing a holographic data set for comprehensively describing the running state of the line; The fault feature extraction module is used for constructing a multi-type fault feature template library and extracting feature data related to various faults from the holographic fusion data by adopting a mixed feature extraction method combining deep learning and traditional signal processing; The holographic diagnosis analysis module integrates an analysis model based on a fault mechanism, a machine learning diagnosis model based on deep learning and a rule reasoning model based on expert experience, and performs full-dimensional and multi-view holographic diagnosis of the fault of the low-voltage distribution line through cooperative linkage of the three models; the accurate positioning calculation module adopts a positioning strategy fused by multiple algorit