CN-122022195-A - Intelligent management method and system for full service life of power distribution automation terminal

CN122022195ACN 122022195 ACN122022195 ACN 122022195ACN-122022195-A

Abstract

The application discloses a full-life intelligent management method and a full-life intelligent management system of a power distribution automation terminal, wherein the method comprises the steps of constructing equipment object fingerprints through equipment ledger data, operation data and parameter data, evaluating equipment object continuity by combining equipment life cycle event sequences and equipment behavior feature sequences, and identifying equipment object offset states; when the offset state is identified, the life cycle account segment division is carried out at the offset detection time point, the attribution correction is carried out on the related historical data, and the equipment life cycle management result is generated. The application can improve the accuracy, the continuity and the traceability of the whole service life management of the power distribution automation terminal.

Inventors

Lin Kongzhou
ZHOU JUN
CHEN WEIWU
WANG DONGJU

Assignees

福建先德能源科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. The intelligent management method for the whole service life of the power distribution automation terminal is characterized by comprising the following steps of: constructing an equipment object fingerprint based on equipment ledger data, equipment operation data and equipment parameter data, wherein the equipment object fingerprint at least comprises an identity feature vector, a position feature vector, a parameter feature vector and a behavior feature vector; Generating a time-ordered sequence of device lifecycle events based on the device operational record; calculating equipment behavior indexes according to a preset statistical window based on the equipment operation data, and generating an equipment behavior feature sequence; inputting the device object fingerprint, the device life cycle event sequence and the device behavior feature sequence into a device object continuity evaluation model to obtain device object continuity confidence; When the equipment object offset state is identified, account segment division is carried out on the equipment life cycle by an offset detection time point to form a life cycle account segment at least comprising an offset pre-account segment and an offset post-account segment; And performing historical data attribution correction on the running data, the operation and maintenance records and the fault records associated with the offset post-account segment based on the fingerprint similarity of the equipment object, and generating an equipment lifecycle management result containing equipment object offset identification and lifecycle account segment information.
2. The method of claim 1, wherein the step of determining the position of the substrate comprises, The equipment account data at least comprises an equipment number, an equipment model number, a manufacturer, a communication identifier, a firmware version number, a delivery date, an installation date, a line, a transformer substation and an equipment running state, the equipment running data at least comprises a data time stamp, three-phase current, three-phase voltage, zero-sequence current, a switching state and a communication state, the equipment parameter data at least comprises a parameter version number, a parameter effective time, an overcurrent protection fixed value, a ground protection fixed value and a communication sampling period, and the equipment operation record at least comprises an operation event type, an event time, an operator and an operation description.
3. The method of claim 2, wherein the step of determining the position of the substrate comprises, The identity feature vector is constructed according to the equipment model number, the manufacturer code, the communication identifier and the firmware version number, the position feature vector is constructed according to the number of a transformer station to which the position feature vector belongs and the number of a line to which the position feature vector belongs, the parameter feature vector is constructed according to an overcurrent protection fixed value, a ground protection fixed value, a sampling period and a communication period, and the behavior feature vector is constructed according to average load current, maximum current, fault action frequency and communication upper line rate obtained through statistics of equipment operation data.
4. The method of claim 3, wherein the step of, The equipment life cycle event sequence consists of an equipment operation event, an equipment maintenance event, a communication identification replacement event, a main board replacement event, an equipment abnormal event, an equipment retired event and an equipment re-used event, wherein when the equipment life cycle event sequence is generated, the event time and the event parameters corresponding to the events are written into the life cycle event table, and the events corresponding to the same equipment number are ordered according to the event time.
5. The method of claim 4, wherein the step of determining the position of the first electrode is performed, And in each time window, calculating average load current and maximum load current according to current records, calculating fault occurrence times according to fault action records, calculating communication online rate according to the online record number and the total record number in communication state records, and arranging corresponding statistical results of each time window according to time sequence to form the equipment behavior feature sequence.
6. The method of claim 5, wherein the step of determining the position of the probe is performed, The device object continuity evaluation model comprises an input layer, a feature embedding layer, a graph convolution layer, a time sequence layer and an output layer, wherein the input layer receives the device object fingerprint, the device life cycle event sequence and the device behavior feature sequence, the feature embedding layer carries out vectorization processing on discrete features in the device object fingerprint, the graph convolution layer carries out association feature extraction based on a device topological relation, the time sequence layer carries out time sequence feature extraction according to the device behavior feature sequence, and the output layer outputs the device object continuity confidence coefficient.
7. The method of claim 6, wherein the step of providing the first layer comprises, When the equipment object offset state is identified, comparing the equipment object continuity confidence with a preset continuity threshold; when the device object continuity confidence is smaller than the continuity threshold, marking the corresponding device as an object offset state, and determining a data time point for generating the mark as an offset detection time point; and when the device object continuity confidence is not smaller than the continuity threshold, keeping the current device life cycle account segment continuous.
8. The method of claim 7, wherein the step of determining the position of the probe is performed, When the device life cycle account segment division is executed, the life cycle record corresponding to the same device number is split into a pre-deviation account segment and a post-deviation account segment by taking the deviation detection time point as a boundary, device operation data, device operation and maintenance records, fault records and parameter records in the corresponding time range are respectively associated with each life cycle account segment, account segment level statistical relations are respectively established, and the subsequent historical data attribution correction is conducted on account segment level data sets.
9. The method of claim 8, wherein the step of determining the position of the first electrode is performed, When the historical data attribution correction is executed, the similarity between the historical data attribution correction and the device object fingerprint corresponding to the offset pre-account section and the device object fingerprint corresponding to the offset post-account section is calculated respectively according to the operation data, the operation and maintenance records and the fault records corresponding to the offset post-account section, each data record is re-associated to the corresponding life cycle account section according to the similarity comparison result, and account section level fault statistics and account section level operation statistics are rebuilt according to the corrected association result.
10. The utility model provides a full life intelligent management system of distribution automation terminal which characterized in that includes: The device object fingerprint construction module is used for constructing device object fingerprints based on device account data, device operation data and device parameter data, wherein the device object fingerprints at least comprise identity feature vectors, position feature vectors, parameter feature vectors and behavior feature vectors; The life cycle event sequence generating module is used for generating a time-sequence arranged life cycle event sequence of the equipment based on the equipment operation and maintenance record; the behavior feature sequence generation module is used for calculating equipment behavior indexes according to a preset statistical window based on the equipment operation data to generate an equipment behavior feature sequence; The object continuity evaluation module is used for inputting the device object fingerprint, the device life cycle event sequence and the device behavior feature sequence into a device object continuity evaluation model so as to obtain device object continuity confidence; When the equipment object offset state is identified, the equipment life cycle is divided by an offset detection time point to form a life cycle account segment at least comprising an offset pre-account segment and an offset post-account segment; And the historical data attribution correction module is used for carrying out historical data attribution correction on the operation data, the operation and maintenance records and the fault records associated with the offset post-account segment based on the fingerprint similarity of the equipment object, and generating an equipment lifecycle management result containing equipment object offset identification and lifecycle account segment information.

Description

Intelligent management method and system for full service life of power distribution automation terminal Technical Field The application relates to the technical field of power distribution automation equipment management, in particular to a full-life intelligent management method and system of a power distribution automation terminal. Background The distribution automation terminal is an important basic device for operation monitoring, fault positioning, protection control and operation and maintenance management of the distribution network, and the operation state, parameter configuration, installation position and operation and maintenance process information of the distribution network directly influence the lean management level of the distribution network. In the prior art, management of a power distribution automation terminal generally takes equipment numbers or account records as main indexes, and records and maintains links such as equipment operation, maintenance, parameter change, abnormal movement, returning operation and re-service. Meanwhile, operation data such as current, voltage, switch state, communication state, protection action and the like, and operation and maintenance records such as overhaul, replacement, parameter adjustment and the like are continuously generated in the operation process of the terminal. Because of the scattered data sources, different formats and different update times, the conventional management mode is difficult to form unified and continuous management for the same terminal throughout the whole life cycle. Particularly, when the terminal is subjected to conditions such as communication identifier replacement, main board replacement, line migration, position adjustment, parameter version switching and the like, although the equipment number may remain unchanged, the identity attribute, the position attribute, the parameter attribute and the behavior characteristic of the equipment object are changed. In the prior art, related historical data is simply accumulated or associated according to the original number, so that data mixing at different stages is easy to occur, and further, equipment life cycle record discontinuity, inaccurate fault statistics, operation analysis distortion and deviation of health evaluation results are caused. On the other hand, for data generated by delayed reporting, entry of the supplemental record or boundary period, the prior art lacks an effective attribution judging and correcting mechanism, and is difficult to accurately distinguish whether the data should correspond to the equipment stage before or after the change. Therefore, a management method capable of realizing continuous identification of equipment objects, segmented management of life cycle and accurate attribution of historical data is provided for solving the problems of inaccurate object identification, insufficient data continuity, disordered attribution of historical data, insufficient staged management capability and the like of a power distribution automation terminal in a whole life cycle, which is a technical problem to be solved in the field. Disclosure of Invention The embodiment of the application provides a full-life intelligent management method and system of a power distribution automation terminal, which are used for at least solving part of technical problems in the related art. According to a first aspect of an embodiment of the present application, there is provided a full-life intelligent management method of a power distribution automation terminal, including: constructing an equipment object fingerprint based on equipment ledger data, equipment operation data and equipment parameter data, wherein the equipment object fingerprint at least comprises an identity feature vector, a position feature vector, a parameter feature vector and a behavior feature vector; Generating a time-ordered sequence of device lifecycle events based on the device operational record; calculating equipment behavior indexes according to a preset statistical window based on the equipment operation data, and generating an equipment behavior feature sequence; inputting the device object fingerprint, the device life cycle event sequence and the device behavior feature sequence into a device object continuity evaluation model to obtain device object continuity confidence; When the equipment object offset state is identified, account segment division is carried out on the equipment life cycle by an offset detection time point to form a life cycle account segment at least comprising an offset pre-account segment and an offset post-account segment; And performing historical data attribution correction on the running data, the operation and maintenance records and the fault records associated with the offset post-account segment based on the fingerprint similarity of the equipment object, and generating an equipment lifecycle management result containing equipment object offset ident