CN-121840888-B - Cross-region synchronous power monitoring system and method based on time sequence database

CN121840888BCN 121840888 BCN121840888 BCN 121840888BCN-121840888-B

Abstract

The invention discloses a transregional synchronous power monitoring system and method based on a time sequence database, and relates to the technical field of data analysis, wherein the method comprises the following steps of acquiring historical abnormal events, storing the historical abnormal events into the time sequence database, extracting first event characteristics and second event characteristics from the historical abnormal events, calculating sensing response time length and recovery processing time length, and taking the sensing response time length and the recovery processing time length as third event characteristics; the method comprises the steps of constructing corresponding event sequence models step by step to obtain complete event sequence model sets of all abnormal events, traversing the sets to screen event pairs, carrying out causal directivity analysis and judgment on causal relations based on the event pairs to generate causal link pair sets, traversing the causal link pair sets, marking and grouping causal link pairs, establishing mapping and storing a knowledge base, marking the result types and average recovery time of each situation element, monitoring situation element abnormality in real time, inquiring the knowledge base to find out common results, calculating estimated maximum recovery time, and early warning if the estimated maximum recovery time exceeds a threshold value.

Inventors

ZENG TIJIAN
ZHU JIANG
LUO YU
DU ZEXIN
LI YUANJUN
XIE ZHIQI
TANG XIAOBO
LI LIN
SU QIAN

Assignees

贵州乌江水电开发有限责任公司

Dates

Publication Date: 20260508
Application Date: 20260313

Claims (10)

1. The cross-region synchronous power monitoring method based on the time sequence database is characterized by comprising the following steps of: The method comprises the steps of acquiring historical abnormal events, storing the historical abnormal events in a time sequence database, extracting first event characteristics and second event characteristics from the historical abnormal events, calculating a perception response time and a recovery processing time based on a time stamp, and taking the perception response time and the recovery processing time as third event characteristics, wherein the first event characteristics represent self attributes of the abnormal events, including fault sources, event types and electrical connection areas or equipment directly associated when the events occur; based on the first, second and third event characteristics corresponding to each historical abnormal event, constructing corresponding event sequence models step by step to obtain a complete event sequence model set of all abnormal events; Traversing the complete event sequence model set, performing time proximity screening and space association screening on each event sequence model serving as a main event, screening event pairs consisting of the main event and candidate events, carrying out causal directivity analysis based on the event pairs to judge causal relations, collecting all the judged causal relations in a link pair mode with the main event as a cause and the candidate event as a result, and generating a causal link pair set; Based on the causal link pair set, marking and grouping associated causal link pairs which are the same in event type and have electrical correlation in influence range, extracting a situation element set and a deviation state index set from the group, establishing a mapping, storing the mapping into a knowledge base, marking the result type and average recovery time of each situation element, monitoring situation element abnormality in real time, inquiring the knowledge base to find out common results, calculating estimated maximum recovery time, and early warning if the maximum recovery time exceeds a threshold value.
2. The method for monitoring the cross-region synchronous power based on the time sequence database according to claim 1, wherein the method comprises the steps of acquiring historical abnormal events and storing the historical abnormal events in the time sequence database, extracting first event features and second event features from the historical abnormal events, calculating a perception response duration and a recovery treatment duration based on a time stamp, and taking the perception response duration and the recovery treatment duration as third event features, and specifically comprises the following steps: respectively establishing connection with data sources of all areas to be monitored, acquiring historical abnormal events of power monitoring equipment of all areas to be monitored, and storing based on a time sequence database, wherein the data sources of the areas to be monitored comprise a SCADA system, a protection information management system, a fault recorder and a sequence event record SOE; Extracting first event characteristics and second event characteristics of each historical abnormal event from a time sequence database, wherein the first event characteristics represent self attributes of the abnormal event, including fault sources, event types and directly related electric connection areas or equipment when the event occurs, the second event characteristics represent first observable influences on a power grid monitoring system before any corrective operation is implemented, the first observable influences comprise electric quantity indexes, equipment state indexes and system service indexes based on operation index states, the electric quantity indexes comprise power or current anomalies of related electric equipment, and voltage anomalies of nodes, the nodes represent common connection points such as buses in a power grid, the equipment state indexes comprise switch deflection, protection device action signals and equipment alarm signals, and the system service indexes comprise SCADA or EMS system data refreshing states and instruction transmission timeliness; based on the time stamp, calculating the time sequence performance characteristics of each abnormal event, including a perception response time length and a recovery treatment time length, wherein the specific calculation steps are as follows: based on the time length from the starting time of the initial event of the fault source to the ending time of capturing and presenting the corresponding second event characteristic, recording as a sensing response time length; Recording a recovery treatment duration based on a duration from a moment when the corresponding second event characteristic is presented to a moment when the electric quantity index, the equipment state index and the system service index are recovered to be within a preset safe and stable threshold range through automatic or manual operation; And taking the calculated sensing response time and the recovery processing time as third event characteristics of each abnormal event.
3. The method for monitoring the cross-region synchronous power based on the time sequence database according to claim 2, wherein the step-by-step construction of the corresponding event sequence model based on the first, second and third event characteristics corresponding to each historical abnormal event is performed to obtain a complete event sequence model set of all abnormal events, and the specific steps include: For each historical abnormal event, connecting a first event feature and a second event feature of the historical abnormal event through a directed edge to generate a basic event sequence model of the event, wherein the directed direction of the directed edge is directed by the first event feature to the second event feature; Connecting the second event feature and the third event feature through a directed edge based on the generated basic event sequence model of the event to generate a complete event sequence model, wherein the pointing direction of the directed edge points to the third event feature from the second event feature; And adding each generated complete event sequence model into the event sequence model set to form a complete event sequence model set covering all historical abnormal events.
4. The method for monitoring the cross-region synchronous power based on the time sequence database according to claim 3, wherein the method comprises the steps of traversing a complete event sequence model set, performing time proximity screening and space association screening on each event sequence model serving as a main event, screening event pairs consisting of the main event and candidate events, judging causal relations based on causal directivity analysis, collecting all the judged causal relations in the form of link pairs taking the main event as a cause and the candidate event as a result, and generating a causal link pair set, and the method comprises the following specific steps of: traversing the complete event sequence model set, recording each event sequence model as a main event M, and executing time proximity screening and space association screening, wherein the specific operation is as follows: searching all other event sequence models which are different from the initial timestamp of the master event M in a preset time window in a time sequence database, and marking the event sequence models as candidate events N; The space association screening is to further screen out the event with intersection or electric connection relation between the influence range in the first event characteristic and the influence range of the main event M from the candidate event N; And carrying out causal directivity analysis on event pairs consisting of the screened main event M and the candidate event N, wherein the method comprises the following specific steps of: Performing time sequence judgment, comparing the initial time stamp of the main event M with the initial time stamp of the candidate event N, and if the initial time of the main event M is earlier than the initial time of the candidate event N, considering that a time sequence premise exists from the main event M to the candidate event N; logic determination is carried out, whether a second event feature of the main event M, namely the influence caused by the main event M, provides a direct reason or a necessary condition for the occurrence of the candidate event N or not is analyzed; if the time sequence premise and the logic possibility are met at the same time, judging that a causal relationship pointed to the candidate event N by the master event M exists; the causal relationship obtained by judging each event pair is expressed in the form of a link pair (M, N), wherein M is a cause and N is a result; All such causal link pairs are assembled, generating a causal link pair set.
5. The method for monitoring the transregional synchronous power based on the time sequence database according to claim 4, wherein the method is characterized by comprising the steps of marking and grouping associated causal link pairs which are electrically associated due to the same event type and the influence range based on causal link pair sets, extracting a situation element set and a deviation state index set from the group, establishing mapping and storing the mapping and the deviation state index set into a knowledge base, marking the result types and average recovery time of each situation element, monitoring situation element abnormality in real time, inquiring the knowledge base to find out common results, calculating estimated maximum recovery time, and early warning if the estimated maximum recovery time exceeds a threshold value, wherein the method comprises the following specific steps: Traversing a causal link pair set, and for each causal link pair CP1 and CP2, wherein in the causal link pair CP1, the causal event is M1, the causal event is N1, and in the causal link pair CP2, the causal event is M2, and the causal event is N2; if the event types of the causal event M1 and the causal event M2 are the same and the influence ranges of the causal event M1 and the causal event M2 have an electrical connection relationship or a topological relationship, marking the causal link pair CP1 and the causal link pair CP2 as a related causal link pair; Grouping all causal link pairs having the same event type and electrical association to form an associated causal link group; For each associated causal chain group, analyzing all causal events in the group, combining historical operation data, extracting common preconditions including power grid structural features, operation state features before the occurrence of the events and environmental condition features, and forming a situation element set F; Establishing mapping relation between the situation element set F and the deviation state index set U; circularly executing the steps, processing all the associated causal chain pairs, generating mapping pairs and storing the mapping pairs into a situation mapping knowledge base; Marking all result event types historically caused by each situation element f on the basis of the generated situation mapping knowledge base, and recording average recovery treatment duration corresponding to each situation element f; The state of each situation element in the network operation process is captured and monitored in real time, and when one or more than one situation element is in an abnormal state, a situation mapping knowledge base is inquired, and the common result event types caused by all activated situation elements are found; Taking the maximum value of the historical average recovery treatment duration corresponding to all activated situation elements as the estimated maximum recovery time, and immediately giving an early warning to the relevant area when the estimated maximum recovery time exceeds a preset acceptable interruption time threshold.
6. A transregional synchronous power monitoring system based on a time sequence database is applied to the transregional synchronous power monitoring method based on the time sequence database according to any one of claims 1-5, and is characterized by comprising a historical abnormal event processing module, an event sequence model construction module, a causal analysis module, a causal link grouping module and an early warning module, wherein the historical abnormal event processing module is used for acquiring historical abnormal events and storing the historical abnormal events into the time sequence database, extracting first event characteristics and second event characteristics from the historical abnormal events, calculating sensing response time and recovery processing time and taking the perceived response time as third event characteristics, the event sequence model construction module is used for constructing corresponding event sequence models step by step to obtain a complete event sequence model set of all abnormal events, the causal analysis module is used for traversing the set to select event pairs, carrying out causal directivity analysis and judgment on the basis of the event sequence model to generate a causal link pair set, the causal link grouping module is used for traversing the causal link pair set, marking the causal link pair and grouping, establishing a mapping knowledge base, marking the result type and average recovery time of each situation element, and the early warning module is used for searching for real-time situation elements, finding out the corresponding event sequence models, calculating the most common recovery time exceeds a threshold value.
7. The transregional synchronous power monitoring system based on the time sequence database, according to claim 6, the historical abnormal event processing module comprises a multi-source data access unit, a time sequence data storage unit, an event feature extraction unit and a time sequence efficiency calculation unit, wherein the multi-source data access unit is used for establishing stable connection with a data source of each region to be monitored to obtain historical abnormal events of the power monitoring equipment, the time sequence data storage unit is used for sequencing the obtained historical abnormal events according to time stamps and storing the time sequence data to the time sequence database, the event feature extraction unit is used for extracting first and second event features of each historical abnormal event from the time sequence database, and the time sequence efficiency calculation unit is used for calculating third event features based on the time stamps.
8. The system for monitoring the power of the transregional synchronization based on the time sequence database, which is characterized in that the event sequence model building module comprises a basic event sequence model generating unit, a complete event sequence model generating unit and an event sequence model set management unit, wherein the basic event sequence model generating unit is used for generating a basic model by connecting a first event feature and a second event feature through a directed edge for each historical abnormal event, the complete event sequence model generating unit is used for completing event full life cycle logic by connecting the second event feature and a third event feature through the directed edge on the basis of the basic model, and the event sequence model set management unit is used for summarizing all complete event sequence models to form a complete event sequence model set.
9. The system for monitoring the power of the transregional synchronization based on the time sequence database according to claim 8, wherein the causal analysis module comprises a model set traversing unit, a time proximity screening unit, a spatial correlation screening unit, a causal directivity analysis unit and a causal link pair set generating unit, wherein the model set traversing unit is used for accessing each model in a complete event sequence model set one by one to define the model as a main event, the time proximity screening unit is used for screening candidate events, the spatial correlation screening unit is used for screening events with an intersection or an electric connection between an influence range and the main event from the candidate events, the causal directivity analysis unit is used for verifying causal relations on event pairs consisting of the main event and the candidate events through time sequence judgment and logic judgment, and the causal link pair set generating unit is used for summarizing the event pairs judged to be causal relations to generate a causal link pair set.
10. The time-series database-based cross-region synchronous power monitoring system of claim 9, wherein the association causal chain grouping module comprises an association causal chain pair marking unit, an extraction unit, a mapping establishment unit, a recovery time length marking unit and an early warning unit, wherein the association causal chain pair marking unit is used for marking link pairs which are electrically associated due to the same event type and have the same influence range and are classified into the same association causal chain group, the extraction unit is used for extracting a situation element set and a deviation state index set for each association causal chain group, the mapping establishment unit is used for establishing a mapping relation, storing the established mapping relation into a knowledge base, and the recovery time length marking unit is used for marking the type of a result event which is historically caused for each situation element in the knowledge base and calculating and storing the corresponding average recovery treatment time length.

Description

Cross-region synchronous power monitoring system and method based on time sequence database Technical Field The invention relates to the technical field of data analysis, in particular to a system and a method for cross-region synchronous power monitoring based on a time sequence database. Background As the power system develops to the cross-regional networking and large-scale new energy grid connection directions, cross-regional power allocation modes such as 'western electric east delivery', 'north-south mutual supply' and the like are mature, the safe and stable operation of the cross-regional power grid becomes the core requirement for guaranteeing energy supply, and the power monitoring is used as a key link of power grid operation and maintenance, and needs to process massive operation data with time stamps in real time, so that the data synchronism, the abnormal event analysis accuracy and the cross-regional association early warning capability of the power grid are influenced, and the power grid fault handling efficiency and the overall power supply reliability are influenced. However, when dealing with multi-source time sequence data integration, abnormal event full-dimension analysis and cross-region causal relation early warning, the traditional cross-region power monitoring method often faces the problems that firstly, multi-source data integration and synchronization difficulty are high, monitoring data sources of a cross-region power grid are distributed and have heterogeneous formats, a single-region independent storage mode is adopted in the traditional monitoring, a unified storage and cross-region data synchronization mechanism based on a time sequence database is lacked, data time stamps are inconsistent, unified analysis of the cross-region is difficult to support, cross-region causal relation and early warning capability are weak, the traditional monitoring is limited to event analysis in a single region, electrical topological relation of the cross-region power grid is not considered, cross-region event pairs which are adjacent in time and are spatially related cannot be screened to judge causal relation, situation element mapping and recovery time length estimation based on a historical event rule are lacked, and early warning is only triggered by a single abnormal index, the problems of false report, missing report or early warning time lag occur easily, and operation and maintenance requirements of the cross-region power grid are difficult to meet. Disclosure of Invention The invention aims to provide a transregional synchronous power monitoring system and method based on a time sequence database, so as to solve the problems in the prior art. In order to achieve the above purpose, the invention provides a method for monitoring cross-region synchronous power based on a time sequence database, which comprises the following steps: The method comprises the steps of acquiring historical abnormal events, storing the historical abnormal events in a time sequence database, extracting first event characteristics and second event characteristics from the historical abnormal events, calculating a perception response time and a recovery processing time based on a time stamp, and taking the perception response time and the recovery processing time as third event characteristics, wherein the first event characteristics represent self attributes of the abnormal events, including fault sources, event types and electrical connection areas or equipment directly associated when the events occur; based on the first, second and third event characteristics corresponding to each historical abnormal event, constructing corresponding event sequence models step by step to obtain a complete event sequence model set of all abnormal events; Traversing the complete event sequence model set, performing time proximity screening and space association screening on each event sequence model serving as a main event, screening event pairs consisting of the main event and candidate events, carrying out causal directivity analysis based on the event pairs to judge causal relations, collecting all the judged causal relations in a link pair mode with the main event as a cause and the candidate event as a result, and generating a causal link pair set; Based on the causal link pair set, marking and grouping associated causal link pairs which are the same in event type and have electrical correlation in influence range, extracting a situation element set and a deviation state index set from the group, establishing a mapping, storing the mapping into a knowledge base, marking the result type and average recovery time of each situation element, monitoring situation element abnormality in real time, inquiring the knowledge base to find out common results, calculating estimated maximum recovery time, and early warning if the maximum recovery time exceeds a threshold value. The method comprises the steps of obtaining historical ab