CN-121979717-A - Fortune dimension analysis system and method

CN121979717ACN 121979717 ACN121979717 ACN 121979717ACN-121979717-A

Abstract

The invention relates to the technical field of operation and maintenance management, in particular to a system and a method for analyzing operation and maintenance data. The method comprises the steps of collecting operation and maintenance event data of a plurality of source systems, carrying out standardized processing, introducing double time dimensions of event time and warehousing time, calculating water level lines of the source systems based on delay feature distribution, generating a global water level line, merging the multi-source operation and maintenance events into an evolutionary transient track and a stable freezing final state track under the constraint of the water level line, realizing smooth freezing of the tracks through a final window mechanism, carrying out controllable processing on the delay event by adopting a mode of combining version evolution and patch tracks, and constructing an operation and maintenance analysis model covering response speed, solving quality and compliance on the basis, so as to generate a comprehensive operation and maintenance data report combining real time and offline. The invention can improve the stability, consistency and interpretability of operation and maintenance track construction and analysis results under the multi-source asynchronous and delay data scene, and has higher engineering application value.

Inventors

LIANG YUGANG

Assignees

上海速擎软件有限公司

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. An operation and maintenance data analysis method applied to an operation and maintenance management system, wherein the operation and maintenance management system comprises a plurality of source systems, and the source systems are used for generating operation and maintenance event data, and the method comprises the following steps of: acquiring operation and maintenance event data of a plurality of source systems, carrying out standardized processing on the operation and maintenance event data, and generating a unified event structure, wherein each event comprises event time and warehouse-in time; According to the warehousing time and the event time, counting delay characteristic distribution of each source system, and calculating water lines of each source system based on the delay characteristic distribution to generate a global water line; merging the operation and maintenance event data into an operation and maintenance track based on the event time and the global water line, wherein the operation and maintenance track comprises a transient track and a final track; Judging whether the transient track meets the termination condition according to the global water line and a preset termination window parameter, converting the transient track meeting the termination condition into a final track and freezing; and constructing an operation and maintenance analysis model based on the transient track and the final state track, and generating a comprehensive operation and maintenance data report through the operation and maintenance analysis model, wherein the operation and maintenance analysis model comprises response speed analysis, solution quality analysis and compliance analysis.
2. The method of claim 1, wherein counting delay profile of each source system based on the time of entry and the time of event, calculating water lines of each source system based on the delay profile, and generating global water lines, comprises: Calculating the delay amount of each event in each source system according to each source system, wherein the delay amount is the difference value between the warehousing time and the event time; based on a preset sliding window, counting the distribution parameters of the delay amount, and generating delay characteristic distribution of the source system; Calculating delay tolerance thresholds of all source systems according to the delay characteristic distribution, wherein the delay tolerance thresholds are delay amounts corresponding to preset high-resolution sites in the delay characteristic distribution; Subtracting the delay tolerance threshold from the maximum event time according to the delay tolerance threshold and the maximum event time recorded by each source system to obtain a water line of each source system; and taking the minimum value of each source system water line to generate a global water line.
3. A method of analyzing a running data according to claim 2, wherein the generating a delay profile of the source system based on a distribution parameter of the delay amount of a preset sliding window includes: Counting the fractional distribution of the delay amount in the sliding window for each event type; according to the quantile distribution, extracting delay amounts corresponding to preset quantiles as distribution parameters; a delay profile of the source system is generated based on the distribution parameters for each event type.
4. The operation and maintenance data analysis method according to claim 1, wherein merging the operation and maintenance event data into an operation and maintenance track based on the event time and the global water line comprises: A stable key and a version number are distributed for each operation and maintenance track, a state identification of the operation and maintenance track and a coverage event time range are maintained, and the coverage event time range comprises the maximum event time and the minimum event time of the operation and maintenance track; when a new operation and maintenance event arrives, searching candidate transient tracks according to the event time and the associated attribute of the new operation and maintenance event; calculating the association degree of the new operation and maintenance event and each candidate transient track; Merging the new operation and maintenance event to a transient track with highest association degree and exceeding a preset association degree threshold value, and updating a coverage event time range of the transient track; if the transient track meeting the condition does not exist, a new transient track is created, and the new operation and maintenance event is classified into the new transient track.
5. The method of claim 4, wherein merging the new operation and maintenance event into a transient track with highest association degree and exceeding a preset association degree threshold value, and updating a coverage event time range of the transient track comprises: comparing the event time of the new operation and maintenance event with the current maximum event time and minimum event time of the transient track; if the event time of the new operation and maintenance event is greater than the maximum event time, updating the maximum event time of the transient track to be the event time of the new operation and maintenance event; If the event time of the new operation and maintenance event is smaller than the minimum event time, updating the minimum event time of the transient track to be the event time of the new operation and maintenance event; The merge operation is recorded in a merge log, which includes an operation type, a trigger condition, an affected track identification, an event identification, an operation time, and a corresponding global water line.
6. The method for analyzing operation and maintenance data according to claim 1, wherein determining whether the transient track satisfies a termination condition according to the global water line and a preset termination window parameter, converting the transient track satisfying the termination condition into a final track, and freezing the final track, comprises: calculating the difference value between the global water line and the termination window parameter to obtain a termination time threshold; Obtaining the maximum event time of the transient track, and judging whether the maximum event time is smaller than the termination time threshold; and if the maximum event time of the transient track is smaller than the ending time threshold, updating the state identification of the transient track to a final state, and freezing the operation and maintenance track.
7. A method of operation and data analysis according to claim 6, further comprising: When a delay event associated with a final track is received, extracting a delay event identifier of the delay event and a merging event identifier of a merged event in the final track; Judging whether the delay event identification of the delay event is completely matched with the merging event identification of at least one event in the final state track; If the final track is completely matched, generating a new version of the final track, classifying the delay event into the new version, and recording version migration information in a merging log; If the final track is not completely matched, calculating the association strength according to the coincidence ratio of the resource identifier and the service identifier, and generating a new version of the final track when the association strength reaches a preset strong association threshold value, and classifying the delay event into the new version; And when the association strength does not reach the strong association threshold, creating a patch track for the delay event, and establishing a reference relation between the patch track and the final track.
8. The method of claim 1, wherein constructing an operation and maintenance analysis model based on the transient trajectory and the final trajectory comprises: Extracting time sequence features of the transient state track and the final state track, wherein the time sequence features comprise event time sequences, state transition time points and track duration; Calculating a response speed index based on the time sequence characteristics, wherein the response speed index comprises a first response time and a stage response interval, the first response time is the difference between the event time of the first response event and the event time of the trigger event in the operation and maintenance track, and the stage response interval is the time difference between adjacent response events; Extracting a solution identification event in the operation and maintenance track, and calculating a solution quality index according to the event type of the solution identification event and the regression identification of a subsequent event, wherein the solution quality index comprises a primary solution rate and the regression times of the problem; Calculating a compliance index according to the event time of the operation and maintenance track and a preset compliance time reference, wherein the compliance index comprises the number of overtime events and the compliance completion rate; And carrying out weighted fusion on the response speed index, the solution quality index and the compliance index to generate an operation and maintenance analysis model.
9. The method of claim 8, wherein generating a comprehensive operational dimension data report from the operational dimension analysis model comprises: Extracting all transient tracks, adding a version number and a replenishable mark for each transient track, and generating a real-time track data set; Inputting the real-time track data set into the operation and maintenance analysis model, calculating a real-time response speed index, a real-time solution quality index and a real-time compliance index, and generating a real-time operation and maintenance data report; Extracting all final track, obtaining frozen version of the final track and data snapshot identification, and generating final track data set; inputting the final state track data set into the operation and maintenance analysis model, calculating a final state response speed index, a final state solution quality index and a final state compliance index, and generating an offline operation and maintenance data report; And merging the real-time operation data report and the offline operation data report to generate a comprehensive operation data report.
10. A fortune dimension analysis system for implementing a fortune dimension analysis method according to any one of claims 1-9, wherein the system comprises a data acquisition module, a water line calculation module, a track construction module, a track freezing module, an analysis model construction module, and a report generation module, wherein: The data acquisition module is used for acquiring operation and maintenance event data of a plurality of source systems, carrying out standardized processing on the operation and maintenance event data and generating a unified event structure, wherein each event comprises event time and warehouse-in time; The water level line calculation module is used for counting delay characteristic distribution of each source system according to the warehouse-in time and the event time, calculating water level lines of each source system based on the delay characteristic distribution and generating a global water level line; The track construction module is used for merging the operation and maintenance event data into an operation and maintenance track based on the event time and the global water line, wherein the operation and maintenance track comprises a transient track and a final track; The track freezing module is used for judging whether the transient track meets the termination condition according to the global water line and a preset termination window parameter, converting the transient track meeting the termination condition into a final track and freezing; the analysis model construction module is used for constructing an operation and maintenance analysis model based on the transient track and the final state track; and the report generation module is used for generating a comprehensive operation and maintenance data report through the operation and maintenance analysis model.

Description

Fortune dimension analysis system and method Technical Field The invention relates to the technical field of operation and maintenance management, in particular to a system and a method for analyzing operation and maintenance data. Background As enterprise informatization and digital operation and maintenance levels increase, the size and complexity of operation and maintenance management data continues to increase. In actual operation and maintenance, the systems such as alarm monitoring, work order disposal and the like independently operate, and different event information is recorded respectively. These event data differ significantly in time, structure, and manner of generation, but business semantics are often directed to the same or the same class of operation and maintenance handling behavior. In order to realize the integral analysis of the operation and maintenance process, etc., the multi-source discrete event association is integrated into a complete operation and maintenance track. In the prior art, the analysis of the operation and maintenance data adopts a streaming or batch processing mode to merge events. The stream processing focuses on real-time performance, supports real-time bulletin boards and instant early warning, and is used for post-hoc analysis and periodic reporting by batch processing depending on data integrity. However, in practical application, the multi-source operation and maintenance system has problems such as event delay, e.g. work order creation delay, change approval memory scaling library delay, etc. Delaying data entry into the analysis system can affect existing operation and maintenance trajectories. Under the existing scheme, the delay event can trigger the repeated operation of the generated track, so that the track structure and the mark are changed frequently, the real-time and offline analysis results are inconsistent, the downstream indexes are affected, and the event attribution change is difficult to interpret. In addition, the prior art lacks a boundary mechanism to limit the influence range of delay data, so that the real-time performance and the final consistency are difficult to balance, and the stability of a running data analysis system is restricted. Therefore, a method and a system for performing stable and controllable merging on operation and maintenance events in a multi-source delay data environment are needed, and the consistency, the interpretability and the auditability of a final operation and maintenance track are ensured while the real-time performance is considered, so that the actual requirements of enterprise operation and maintenance analysis and management are met. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a system and a method for analyzing operation and maintenance data, wherein the method collects the operation and maintenance event data of a plurality of source systems and performs standardized processing, introduces the double time dimension of event time and warehouse-in time, calculates the water level line of each source system based on delay characteristic distribution and generates a global water level line; under the constraint of a water line, merging and constructing a multisource operation and maintenance event into an evolution transient track and a stable freezing final track, realizing smooth freezing of the track through a final window mechanism, performing controllable processing in a version evolution and patch track combination mode aiming at a delayed event, constructing an operation and maintenance analysis model covering response speed, solving quality and compliance on the basis, and generating a real-time and offline combination comprehensive operation and maintenance data report. In order to achieve the above purpose, the present invention provides the following technical solutions: an operation and maintenance data analysis method applied to an operation and maintenance management system, the operation and maintenance management system comprising a plurality of source systems, the source systems being used for generating operation and maintenance event data, the method comprising: acquiring operation and maintenance event data of a plurality of source systems, carrying out standardized processing on the operation and maintenance event data, and generating a unified event structure, wherein each event comprises event time and warehouse-in time; According to the warehousing time and the event time, counting delay characteristic distribution of each source system, and calculating water lines of each source system based on the delay characteristic distribution to generate a global water line; merging the operation and maintenance event data into an operation and maintenance track based on the event time and the global water line, wherein the operation and maintenance track comprises a transient track and a final track; Judging whether the transient track meets the te