Transformer substation switching operation risk prediction method based on big data analysis

Abstract

The invention provides a transformer substation switching operation risk prediction method based on big data analysis, and belongs to the technical field of power system transformer substation operation and maintenance, and the method comprises the steps: firstly collecting operation execution information, equipment operation information, environment influence information, historical operation risk information and other multi-source associated information related to transformer substation switching operation; constructing a one-way association link set between the operation and the risk factor; constructing a switching operation risk association network based on the set and a risk evolution path in historical operation risk information; performing risk conduction intensity analysis on nodes and conduction relations in the network to generate risk conduction weight description; and finally, deducing a risk evolution path and outputting a risk prediction result in combination with the weight description and the current real-time execution information. According to the method, the multi-source factor association relationship can be comprehensively considered, the risk is dynamically predicted, and the safety and reliability of the switching operation of the transformer substation are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of power system substation operation and maintenance technology, and more specifically, to a method for predicting the risk of substation switching operations based on big data analysis. Background Technology

[0002] In power systems, substation switching operations are a critical and complex task, involving numerous steps and factors, and playing a vital role in the safe and stable operation of the power system. Traditional risk assessment methods for substation switching operations have many limitations.

[0003] On the one hand, traditional methods often focus only on the impact of a single factor on the risk of switching operations, such as considering only the equipment operating status or the operator's skill level, while ignoring the complex interrelationships between these factors. However, in actual operation, multiple factors such as operational execution, equipment operating conditions, environmental conditions, and historical operational experience intertwine and influence each other, jointly determining the degree of risk in switching operations. The aforementioned single-factor analysis method cannot comprehensively and accurately assess operational risks, easily leading to significant deviations between risk assessment results and actual conditions.

[0004] On the other hand, traditional methods lack effective analysis of the dynamic evolution of risks. During substation switching operations, risks are not static but constantly evolve as the operation progresses and various factors change. Traditional methods struggle to capture these dynamic changes and cannot predict risk trends in advance, thus increasing safety hazards during operations. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for predicting the risk of substation switching operations based on big data analysis, the method comprising: Collect multi-source correlation information related to substation switching operations, including operation execution information, equipment operation information, environmental impact information, and historical operation risk information; Risk factor association links are constructed from multi-source related information to generate a set of one-way association links between operations and risk factors. Each association link in the set of one-way association links corresponds to a mapping relationship between a single operation and a single risk factor. Based on the risk evolution path in the set of unidirectional associated links and historical operation risk information, a switching operation risk association network is constructed. The switching operation risk association network includes a description of the associated link nodes and the transmission relationship between the nodes. A risk transmission intensity analysis is performed on the nodes and transmission relationships in the risk association network of switching operations, and a risk transmission weight description is generated for each associated link. By combining the risk transmission weight description with the real-time execution information of the current switching operation, the risk evolution path in the risk association network of switching operation is deduced, and the risk prediction results of substation switching operation are output.

[0006] Furthermore, embodiments of the present invention also provide a substation switching operation risk prediction system based on big data analysis, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the above-described substation switching operation risk prediction method based on big data analysis by executing the machine-executable instructions.

[0007] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, the processor of a computer device reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the computer device to execute the above-mentioned substation switching operation risk prediction method based on big data analysis.

[0008] Based on the above, multi-source correlation information related to substation switching operations is collected, covering aspects such as operation execution, equipment operation, environmental impact, and historical operational risks. A set of unidirectional correlation links between operations and risk factors is then constructed, clarifying the mapping relationship between each operational step and a single risk factor. This makes the correlation between risk factors and operational steps more intuitive and concrete. The resulting switching operation risk correlation network further reveals the transmission relationships between various risk factors, enabling a comprehensive understanding of the risk propagation path and evolution patterns. Next, risk transmission intensity analysis is performed on the nodes and transmission relationships in the risk correlation network, generating risk transmission weight descriptions for each correlation link. This quantifies the degree of risk transmission between each link. Combining these risk transmission weight descriptions with real-time execution information of the current switching operation, the risk evolution path is extrapolated. This allows for early prediction of substation switching operation risks, providing timely and accurate early warning information for operators, effectively reducing operational risks, improving the safety and reliability of substation switching operations, and ensuring the stable operation of the power system. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the execution flow of the substation switching operation risk prediction method based on big data analysis provided in an embodiment of the present invention.

[0010] Figure 2 This is a schematic diagram of exemplary hardware and software components of the substation switching operation risk prediction system based on big data analysis provided in an embodiment of the present invention. Detailed Implementation

[0011] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating a method for predicting the risk of substation switching operations based on big data analysis, provided in one embodiment of the present invention. The following is a detailed description of this method for predicting the risk of substation switching operations based on big data analysis.

[0012] Step S110: Collect multi-source correlation information related to substation switching operations. The multi-source correlation information includes operation execution information, equipment operation information, environmental impact information, and historical operation risk information.

[0013] In this embodiment, the switching operation of a 10kV line bay in a 110kV substation is taken as the application scenario. The switching operation includes continuous operation steps such as circuit breaker opening, disconnecting switch opening, and grounding switch closing.

[0014] When collecting multi-source related information, operation execution information is obtained through the operation ticket execution record module of the substation monitoring system, including operation task number, operation sequence, planned execution time of each operation step, actual execution time, operator, and supervisor data. Equipment operation information is collected through sensors and intelligent terminals installed on various electrical equipment, covering circuit breaker opening and closing coil current, mechanical characteristic parameters (such as opening time, closing time, and bounce time), auxiliary contact status of disconnecting switches, position signals of grounding switches, and real-time measurement values ​​of current transformers and voltage transformers. The acquisition frequency is set to once per second, and the data storage format adopts the SCL format of the IEC61850 standard. Environmental impact information is obtained through environmental monitoring devices deployed in the substation, including ambient temperature and relative humidity data collected by temperature and humidity sensors, real-time wind speed and direction collected by anemometers, and lightning strike count and lightning current amplitude recorded by lightning monitoring devices. The acquisition time of environmental data is synchronized with the equipment operation information.

[0015] Historical operational risk information is extracted from the substation's safety supervision and management system database. It includes records of risk events that occurred during switching operations of the 10kV line bay and similar bays over the past five years. Each record details the time of the risk event, the corresponding operational steps, the equipment involved, the environmental conditions at the time, the type of risk event (such as misoperation, equipment failure, protection malfunction, etc.), the impact, and the handling measures.

[0016] During the collection of the aforementioned multi-source related information, data anonymization techniques were used to process sensitive data such as operator identity information and device passwords. Specifically, operator names and ID numbers were anonymized by replacing real names with operation numbers, and only the first six and last four digits of the ID number were retained, with the middle digits replaced by asterisks. Sensitive information such as device passwords was encrypted using irreversible encryption algorithms to ensure that sensitive data was not leaked during data transmission and storage.

[0017] Step S120: Construct risk factor association links for multi-source related information, and generate a set of one-way association links between operations and risk factors. Each association link in the set of one-way association links corresponds to the mapping relationship between a single operation link and a single risk factor.

[0018] Step S121: Decompose the operation execution information in the multi-source association information into operation links to obtain multiple continuous operation link units. Each operation link unit corresponds to a specific execution action in the switching operation. Add a time sequence identifier and a device association identifier to each operation link unit.

[0019] In the aforementioned 10kV line bay switching operation scenario, the operation ticket content for the switching operation is extracted from the operation execution information. Based on the operation steps on the operation ticket, the entire switching operation is broken down into multiple consecutive operation units. For example, "opening the 10kV line circuit breaker" is considered the first operation unit, "opening the 10kV line disconnector" is the second operation unit, and "closing the 10kV line grounding switch" is the third operation unit, and so on. Each operation unit corresponds to a specific execution action. A time sequence identifier is added to each operation unit, using the operation sequence number as the time sequence identifier. For example, the time sequence identifier for the first operation unit is "001", the second is "002", and so on. The equipment association identifier is set according to the specific equipment involved in the operation unit. For example, the operation unit of "opening the 10kV line circuit breaker" corresponds to the circuit breaker of the 10kV line. The equipment association identifier is set as the unique equipment number of the circuit breaker in the substation, such as "10kV-L1-B01", where "10kV-L1" represents the 10kV line bay number and "B01" represents the circuit breaker number.

[0020] Step S122: Select risk factors associated with each operation unit from the equipment operation information, environmental impact information and historical operation risk information in the multi-source associated information. First, match abnormal parameter records in the equipment operation information by equipment association identifier, then match the same-period environmental change data in the environmental impact information by time series identifier, and finally match similar operation risk records in the historical operation risk information by the action type of the operation unit to form a risk factor list corresponding to each operation unit.

[0021] Step S1221: Classify and organize the equipment operation information. Divide the equipment operation information into multiple equipment status datasets according to the equipment type. Each equipment status dataset contains the corresponding equipment's operating parameter change information, fault record information, maintenance record information, and parameter abnormality alarm records. Add a unique equipment identifier to each equipment status dataset.

[0022] For the aforementioned 10kV line bay switching operations, the equipment operation information is categorized by equipment type into circuit breaker equipment status datasets, disconnector switch equipment status datasets, and grounding switch equipment status datasets. Taking the circuit breaker equipment status dataset as an example, it includes the circuit breaker's opening and closing coil current variation curve over time (operating parameter variation information), fault records from the past year such as mechanism jamming and auxiliary power supply failures (fault record information), maintenance records including the time, maintenance items, and replaced components of the most recent maintenance (maintenance record information), and parameter anomaly alarm records generated when the opening time exceeds the rated value or the closing coil current abnormally increases. The unique equipment identifier added to this circuit breaker equipment status dataset is consistent with the equipment association identifier added to the operation unit in step S121, namely "10kV-L1-B01".

[0023] Step S1222: Dimensionally decompose the environmental impact information into three dimensions: natural environment information, on-site operation environment information, and external disturbance environment information. Each dimension of environmental data includes data collection time, data value, and data change trend record.

[0024] The collected environmental impact information was broken down into dimensions. The natural environment information dimension includes data such as ambient temperature, relative humidity, wind speed, wind direction, rainfall, and lightning activity. Each data item includes the specific collection time (accurate to the second), the measured data value, and a record of data change trends calculated based on continuously collected data, such as the temperature rise or fall trend over the past hour, and the magnitude of wind speed changes. The on-site operational environment information dimension covers the light intensity, noise level, and dust concentration at the operating site, also including the collection time, data values, and change trend records. The external interference environment information dimension mainly includes data on electromagnetic interference intensity from nearby high-voltage lines and radio communication interference, also including the corresponding collection time, data values, and change trend records.

[0025] Step S1223: Perform structured processing on historical operational risk information, classify historical risk events according to risk event type, extract the associated equipment information, associated environmental information, associated operational information, risk occurrence time, risk impact scope and risk handling result corresponding to each historical risk event, and establish a historical risk event index library.

[0026] Risk events in historical operational risk information are categorized by type, including misoperation risk events, equipment failure risk events, protection device malfunction risk events, and environmental impact risk events. For each historical risk event, its associated equipment information is extracted, such as the 10kV line disconnector switch number involved in a misoperation risk event; associated environmental information, such as the ambient temperature and whether there was thunderstorm weather; associated operational steps, such as the "opening the disconnector switch" operation; the time of risk occurrence, accurate to year, month, day, hour, minute, and second; the scope of risk impact, such as whether it caused a power outage and the number of affected users; and the risk handling results, such as the emergency power-off measures taken and the replacement of faulty equipment. The extracted information is entered into a database to establish a historical risk event index. The index keywords include risk event type, associated equipment information, and associated operational step information to facilitate rapid subsequent querying and matching.

[0027] Step S1224: For each operation unit, match the corresponding equipment status data in the equipment status dataset with its equipment association identifier, analyze the operating parameter change information and parameter abnormality alarm records in the equipment status data, and determine the equipment-related risk factors that may affect the execution of the operation unit. Equipment-related risk factors include parameter abnormality risk, equipment failure risk, and maintenance deficiency risk.

[0028] Taking the operation of "opening a 10kV line circuit breaker" as an example, its equipment association identifier is "10kV-L1-B01". The corresponding circuit breaker equipment status data is matched in the equipment status dataset using this identifier. Analyzing the changes in the circuit breaker's operating parameters reveals that if the trip coil current shows a gradually increasing trend in recent operations, and the tripping time slightly exceeds the rated upper limit in a certain operation, these abnormal changes in operating parameters may lead to an excessively long tripping time in this operation, thus constituting a parameter anomaly risk. Checking the fault record information, if the circuit breaker experienced a trip coil disconnection fault within the past three months, although it has been repaired, there is still a possibility of a similar fault recurring; this is a equipment fault risk. Reviewing the maintenance record information, if the circuit breaker's scheduled maintenance cycle is six months, and the current maintenance period is more than seven months old, exceeding the prescribed maintenance cycle, there may be problems such as mechanical component wear and insufficient lubrication due to lack of maintenance; this is a maintenance deficiency risk. The risks of abnormal parameters, equipment failure, and lack of maintenance mentioned above are summarized as equipment-related risk factors for this operational unit.

[0029] Step S1225: Filter environmental data that overlaps with the execution time of the operation unit from the three dimensions of environmental data, analyze extreme value records in the natural environment information, non-compliance records in the on-site operation environment information, and interference intensity records in the external interference environment information to determine environmental risk factors that may affect the execution of the operation unit. Environmental risk factors include natural environment anomaly risk, non-compliance risk in the operation environment, and external interference risk.

[0030] Suppose the planned execution time for the "opening the 10kV line circuit breaker" operation is from 9:00 AM to 9:05 AM on a certain morning. From the perspective of natural environmental information, if the ambient temperature reaches the highest historical value for the same period in the region, and the relative humidity is below the specified lower limit, it may lead to heatstroke for operators and a decrease in equipment insulation performance; this is a risk of abnormal natural environment. From the perspective of on-site work environment information, if the measured light intensity at the operation site is lower than the minimum light standard specified in the substation safety operation procedures, it will affect the operator's ability to observe equipment status; this is a risk of substandard work environment. From the perspective of external interference environment information, if de-icing operations are detected on a nearby high-voltage line at the same time, generating strong electromagnetic interference, it may interfere with the circuit breaker's control circuit, affecting the normal execution of the opening operation; this is a risk of external interference. The above-mentioned risks of abnormal natural environment, substandard work environment, and external interference are identified as the environmental risk factors corresponding to this operation unit.

[0031] Step S1226: Filter historical risk events from the historical risk event index library whose associated operation information is consistent with the action type of the current operation unit. Extract the risk factors, associated equipment information and associated environmental information corresponding to these historical risk events. Compare the equipment association identifier of the current operation unit with the associated equipment information of the historical risk events. Filter the historical risk factors that match the equipment as historical associated risk factors.

[0032] In the historical risk event index, using "opening the circuit breaker" as the search keyword for related operation information, all historical risk events with the action type of opening the circuit breaker were filtered out. The risk factors corresponding to these historical risk events, such as "circuit breaker tripping without authorization," "control circuit disconnection," and "operating mechanism refusing to operate," were extracted, along with the corresponding associated equipment information (such as circuit breaker number) and associated environmental information. The equipment association identifier "10kV-L1-B01" of the current operation unit was compared with the associated equipment information of historical risk events, filtering out historical risk events with the same associated equipment information as "10kV-L1-B01" or the same model and batch of circuit breakers. For example, if historically, this model of circuit breaker failed to operate during the "opening the circuit breaker" operation due to the tripping of the auxiliary power supply air switch in the control circuit, this risk factor is a historical associated risk factor matching the equipment of the current operation unit.

[0033] Step S1227: Merge equipment-related risk factors, environmental-related risk factors, and historically related risk factors. Compare the consistency of risk factor descriptions, eliminate completely duplicate risk factors, and add risk type identifiers and associated data source identifiers to the retained risk factors. Risk type identifiers are divided into equipment-related risk identifiers, environmental risk identifiers, and historically related risk identifiers. Form a risk factor list corresponding to each operational unit, verify the completeness of the risk factor list, and supplement any missing risk factor types.

[0034] The equipment-related risk factors (parameter anomaly risk, equipment failure risk, maintenance deficiency risk) obtained in step S1224, the environmental risk factors (natural environment anomaly risk, substandard working environment risk, external interference risk) obtained in step S1225, and the historically related risk factors (such as control circuit auxiliary power supply air switch tripping risk) obtained in step S1226 are merged. The descriptions of each risk factor are compared; if there are risk factors with completely identical descriptions, such as "control circuit disconnection" in equipment-related risk factors and "control circuit disconnection" in historically related risk factors, duplicates are removed. A risk type identifier is added to each retained risk factor, such as "equipment-related" for equipment-related risk factors, "environment-related" for environmental-related risk factors, and "historically related" for historically related risk factors. Simultaneously, a data source identifier is added, such as "circuit breaker equipment status dataset - operating parameter change information" for parameter anomaly risk, and "natural environment information dimension - temperature data" for natural environment anomaly risk, etc. After creating the risk factor list for the "opening the 10kV line circuit breaker" operation unit, check whether the list includes risk factors in three aspects: equipment, environment, and historical correlation. If a certain type of risk factor is found to be missing, such as the risk of "hydraulic oil leakage of the operating mechanism" in the historical correlation category, then add the risk factor to the list to ensure the completeness of the risk factor list.

[0035] Step S123: Analyze the triggering relationship between each operation unit and the corresponding risk factors in the risk factor list. By comparing the execution time of the operation unit and the occurrence time of the risk factors in the historical operation risk information, determine the direction attribute of the triggering relationship. The direction attribute is whether the operation unit triggers the risk factor or the risk factor affects the operation unit.

[0036] For each risk factor in the "opening the 10kV line circuit breaker" operation unit and its risk factor list, the triggering relationships are analyzed one by one. Taking the risk factor "operating mechanism failure to operate" as an example, check whether there are records of operating mechanism failure during the execution of this operation unit in the historical operation risk information. If in a certain historical operation, the operator performed the "open the circuit breaker" operation (operation unit execution time), and the circuit breaker failed to open normally afterward, and the inspection found that the operating mechanism was stuck, causing the failure to operate (risk factor occurrence time), then the operation unit execution time comes first, and the risk factor occurrence time comes later, indicating that the execution of the operation unit triggered the occurrence of the risk factor, and the direction attribute is determined to be that the operation unit triggered the risk factor. Taking the risk factor "excessively high ambient temperature" as another example, if the historical record shows that the ambient temperature was already excessively high before a certain "open the circuit breaker" operation (risk factor occurrence time), and the high temperature caused the operator's actions to be slow during the operation, affecting the efficiency of the operation (operation unit execution time comes later), then it indicates that the risk factor affected the execution of the operation unit, and the direction attribute is determined to be that the risk factor affected the operation unit. By performing the above analysis on all risk factors in the risk factor list, the triggering relationship direction attribute between each risk factor and the operational unit is clarified.

[0037] Step S124: Based on the direction attribute of the triggering relationship, construct a one-way association link between each operation link unit and the corresponding risk factor. The one-way association link includes the operation link identifier, risk factor identifier, triggering direction identifier, and association basis record. The association basis record is the matched equipment abnormal parameter record, environmental change data, or historical risk record.

[0038] Based on the triggering relationship direction attribute determined in step S123, a one-way association link is established between the "open 10kV line circuit breaker" operation unit and each risk factor. The operation unit identifier is set as a combination of the operation sequence number and operation content of the operation unit, such as "001-open 10kV line circuit breaker". The risk factor identifier is a unique code for each risk factor, such as "SFXY-001" for parameter abnormality risk and "SBGZ-001" for equipment failure risk. The triggering direction identifier is set according to the direction attribute. When the operation unit triggers a risk factor, the identifier is "C→R" (C represents the operation unit, R represents the risk factor), and when the risk factor affects the operation unit, the identifier is "R→C". The association basis record references the specific matched data in detail. For example, the association basis record for parameter abnormality risk is the specific record number and corresponding parameter value of the circuit breaker equipment status data showing that the centralized tripping time exceeds the rated range; the association basis record for environmental change data is the record number of the environmental monitoring device showing that the temperature and humidity exceeded the standard during the operation period; and the association basis record for historical risk records is the corresponding risk event number and brief description in the historical risk event index database.

[0039] Step S125: Perform a repeatability check on all unidirectional links, compare the consistency of the combination of operation link identifier, risk factor identifier and trigger direction identifier, eliminate unidirectional links with completely identical combinations, and retain unique unidirectional links.

[0040] Iterate through all established one-way links, comparing the operation step identifier, risk factor identifier, and trigger direction identifier of each link as a combination. If two one-way links have the same combination of these three identifiers, for example, both being "001-Open 10kV line circuit breaker", "SFXY-001", and "C→R", they are identified as duplicate links. Analyze the duplicate links, checking for differences in their association basis records. If the association basis records differ but the core risk factors and trigger relationships are consistent, retain only one link and merge the association basis records. If the association basis records point to different risk scenarios, reassess whether it is a duplicate link. After verification, remove all one-way links with identical combinations, ensuring that each retained one-way link is unique.

[0041] Step S126: Sort the unique unidirectional association links retained according to the execution order corresponding to the time sequence identifier of the operation link unit, and mark the source of the association basis record for each unidirectional association link to form a set of unidirectional association links between operations and risk factors.

[0042] The unidirectional links retained after repeatability verification are arranged sequentially according to the execution order corresponding to the time sequence identifiers (such as "001", "002", "003", etc.) of the operation unit. For unidirectional links corresponding to operation units such as "opening the 10kV line circuit breaker" (time sequence identifier "001") and "opening the 10kV line disconnector" (time sequence identifier "002"), they are arranged in the order of "001" and "002". Simultaneously, the specific source of the association record is marked next to each unidirectional link. For example, equipment abnormal parameter records come from "Circuit Breaker Equipment Status Dataset (10kV-L1-B01)", environmental change data comes from "Natural Environment Information Dimension - Temperature and Humidity Sensor (No. TH-001)", and historical risk records come from "Historical Risk Event Index Library - Event No. FX-2022-035", etc. The sorted and marked unidirectional links are integrated to form a set of unidirectional links between operations and risk factors.

[0043] Step S130: Based on the risk evolution path in the unidirectional association link set and historical operation risk information, construct a switching operation risk association network, which includes the association link nodes and the description of the transmission relationship between the nodes.

[0044] Step S131: Extract the risk evolution path corresponding to the risk events that have occurred from the historical operational risk information. Each risk evolution path contains multiple risk factors that are triggered in succession and the corresponding sequence of operational links. At the same time, record the equipment status snapshot and environmental status snapshot when each risk factor and operational link is triggered.

[0045] From the historical operational risk information, a past "disconnect switch being pulled under load" risk event in this 10kV line bay was selected as an example. The risk evolution path corresponding to this risk event was extracted as follows: First, the operator did not strictly follow the operation ticket verification process (operation link unit), resulting in the disconnect switch being mistakenly operated as a circuit breaker (risk factor A); then, the disconnect switch was operated under load (operation link unit), generating an electric arc (risk factor B); next, the electric arc failed to be extinguished in time (risk factor C), leading to a phase-to-phase short circuit (risk factor D); finally, the relay protection device tripped (operation link unit). The above risk evolution path includes the consecutively triggered risk factors A, B, C, and D and their corresponding operation link unit sequences. Simultaneously, snapshots of the equipment status at the time of each risk factor and operation link unit trigger were recorded, such as the position signal and auxiliary contact status of the disconnect switch when risk factor A was triggered, and the secondary current value of the current transformer when risk factor B was triggered; environmental status snapshots, such as the light intensity at the operation site when risk factor A was triggered, and the ambient temperature when risk factor B was triggered, were also recorded.

[0046] Step S132: Take each one-way associated link in the set of one-way associated links as the basic node of the switching operation risk association network, and assign a unique node number to the basic node. The basic node includes the operation link identifier, risk factor identifier, trigger direction identifier and association basis record.

[0047] Each unidirectional associated link in the set of unidirectional associated links formed in step S126 is converted into a basic node of the switching operation risk association network. For example, the unidirectional associated link between the "opening the 10kV line circuit breaker" operation unit and the risk factor "operating mechanism refusing to operate" (trigger direction identifier "C→R") is considered as a basic node. A unique node number is assigned to this basic node, using the numbering rule of "node type prefix + sequence number". For example, if the node prefix for the operation unit triggering the risk factor type is "CR-", and the sequence number starts from 001 and increases sequentially, the basic node number can be set to "CR-001". The basic node contains the operation unit identifier "001-opening the 10kV line circuit breaker", the risk factor identifier "SBGZ-001" (the identifier corresponding to the operating mechanism refusing to operate), the trigger direction identifier "C→R", and the association basis record is the reference information of the equipment abnormal parameter record, environmental change data, or historical risk record corresponding to this unidirectional associated link.

[0048] Step S133: Analyze the connection relationship between continuous operational units and risk factors in the risk evolution path, extract the trigger interval time between adjacent operational units and risk factors in the risk evolution path, and determine the transmission order between basic nodes by combining the changing trends of equipment status snapshots and environmental status snapshots. The transmission order is consistent with the trigger order in the risk evolution path.

[0049] Step S1331: Segment each risk evolution path and divide it into multiple consecutive trigger segments in chronological order. Each trigger segment contains an operational unit, a directly triggered risk factor, and the corresponding trigger timestamps of both.

[0050] The risk evolution path of the aforementioned "disconnecting switch operated under load" risk event is segmented chronologically. The first triggering segment is: the operational unit "operator failed to strictly follow the operation ticket verification process," directly triggering risk factor A "mistakenly operating the disconnecting switch as a circuit breaker." The triggering timestamp for the operational unit is T1, and the occurrence timestamp for risk factor A is T1+Δt1. The second triggering segment is: the operational unit "disconnecting switch operated under load," directly triggering risk factor B "arc generation." The triggering timestamp for the operational unit is T2, and the occurrence timestamp for risk factor B is T2+Δt2. The third triggering segment is: risk factor B "arc generation," directly triggering risk factor C "arc failure to extinguish in time." The occurrence timestamp for risk factor B is T2+Δt2, and the occurrence timestamp for risk factor C is T2+Δt2+Δt3. The fourth trigger segment is: Risk factor C "failure to extinguish the arc in time", which directly triggers risk factor D "phase-to-phase short circuit". The timestamp of risk factor C is T2+Δt2+Δt3, and the timestamp of risk factor D is T2+Δt2+Δt3+Δt4. The fifth trigger segment is: Risk factor D "phase-to-phase short circuit", which triggers the operation unit "relay protection device trips". The timestamp of risk factor D is T2+Δt2+Δt3+Δt4, and the timestamp of the operation unit is T3. Each trigger segment includes an operation unit or risk factor, a directly triggered risk factor or operation unit, and two corresponding trigger timestamps.

[0051] Step S1332: Match each trigger segment with the unidirectional association links in the unidirectional association link set, compare the combination of operation link units and risk factors in the trigger segment, determine the basic node corresponding to each trigger segment, and record the number of the basic node that is successfully matched.

[0052] The combination of the operational step "operator did not strictly follow the operation ticket verification process" in the first trigger segment and risk factor A "mistakenly operating the isolating switch as a circuit breaker" is searched in the unidirectional associated link set. If there is a unidirectional associated link in the unidirectional associated link set with the operational step identifier "operator did not strictly follow the operation ticket verification process", the risk factor identifier "misoperation-001" (assuming the identifier of risk factor A), and the trigger direction identifier "C→R", then the trigger segment is matched to the basic node corresponding to the unidirectional associated link, and the number of the basic node is recorded, such as "CR-005". In the same way, other trigger segments are matched one by one with the unidirectional associated links in the unidirectional associated link set to determine the basic node number corresponding to each trigger segment.

[0053] Step S1333: Track the basic nodes corresponding to adjacent trigger segments in the same risk evolution path, extract the association basis records of the basic nodes corresponding to the previous trigger segment and the association basis records of the basic nodes corresponding to the next trigger segment, analyze the correlation between the two in terms of equipment status parameters and environmental status parameters, and determine the connection relationship between the output end of the previous basic node and the input end of the next basic node.

[0054] Taking the basic nodes (node ​​numbers CR-005 and CR-008) corresponding to the first and second trigger segments in the risk evolution path as examples, the associated basis records of CR-005 are extracted. These records may include the monitoring video time point when the operator failed to verify the operation ticket, the operation record of the operation ticket system, etc. The associated basis records of CR-008 are extracted, which may include the current value during the operation of the disconnecting switch, the action signal of the operating mechanism, etc. The continuity of the operation time in the associated basis records of CR-005 and CR-008 is analyzed, as well as whether the operator error involved in CR-005 directly caused the change in the equipment status parameters of the disconnecting switch in CR-008 under load. If the associated basis records of CR-005 show that the operator failed to verify the equipment number, directly leading to the wrong operation object, and thus causing the disconnecting switch in CR-008 to be operated under load current, it can be determined that there is a direct connection between the output terminal of CR-005 and the input terminal of CR-008, that is, the risk factor output of the previous basic node triggers the operation link unit input of the next basic node.

[0055] Step S1334: Determine the transmission direction between basic nodes based on the connection relationship. The transmission direction is from the basic node that is triggered first to the basic node that is triggered later. At the same time, mark the parameter association type corresponding to the connection relationship. The parameter association type is either device parameter association or environmental parameter association.

[0056] Based on the connection relationship determined in step S1333, the first triggered basic node CR-005 precedes the second triggered basic node CR-008. Therefore, the transmission direction is determined to be from CR-005 to CR-008. Analyzing the parameter correlation in the correlation records of the two, the correlation record of CR-005 mainly involves operational behavior parameters (non-equipment or environmental parameters), while the correlation record of CR-008 involves the current parameters of the disconnecting switch (equipment parameters). Therefore, the parameter correlation type is labeled as equipment parameter correlation. If the correlation records of two adjacent basic nodes involve the transmission of changes in environmental parameters such as ambient temperature and humidity, the parameter correlation type is labeled as environmental parameter correlation.

[0057] Step S1335: Summarize the transmission directions of the basic nodes corresponding to all risk evolution paths, count the frequency of the transmission direction corresponding to each combination of basic nodes, remove contradictory transmission direction descriptions whose transmission direction frequency is lower than a preset ratio, and retain the transmission direction consistent with most risk evolution paths.

[0058] Collect the transmission directions of basic nodes corresponding to the risk evolution paths of all historical risk events. For example, in 10 risk evolution paths, the transmission direction of the basic node combination (CR-005, CR-008) is CR-005→CR-008 in 8 paths, and CR-008→CR-005 (contradictory transmission direction) in 2 paths. A preset ratio is set to 50%, meaning that when the frequency of contradictory transmission directions is less than 50% of the total frequency, the contradictory transmission direction description is removed. In the example above, the frequency of contradictory transmission directions is 20%, which is less than 50%, therefore the transmission direction description of CR-008→CR-005 is removed, and the transmission direction of CR-005→CR-008 is retained. This method is applied to different basic node combinations to ensure that the retained transmission directions are consistent with the majority of risk evolution paths.

[0059] Step S1336: According to the retained transmission direction, and in combination with the parameter association type, organize the sequential triggering logic of all basic nodes, establish a basic node triggering sequence table, and form the transmission order between basic nodes. The transmission order is consistent with the triggering order in the risk evolution path. The basic node triggering sequence table includes node number, preceding node number, succeeding node number and parameter association type.

[0060] Based on the retained transmission direction and parameter association type, the sequential triggering logic of the basic nodes is organized. Taking basic node CR-005 as an example, its preceding node number is empty (indicating that this node is the starting node of a certain transmission path), and its following node number is CR-008, with a parameter association type of device parameter association. The preceding node number of basic node CR-008 is CR-005, and its following node number is CR-012 (assuming the next basic node number), with a parameter association type of environmental parameter association (if its connection with CR-012 involves environmental parameters). The above information of all basic nodes is entered into the basic node triggering sequence table. This basic node triggering sequence table clearly shows the node number, preceding node number (i.e., the number of the previously triggered basic node), following node number (i.e., the number of the next triggered basic node), and parameter association type of each basic node, thus forming the transmission order between basic nodes. This transmission order is completely consistent with the triggering order in the risk evolution path.

[0061] Step S134: Establish a transmission connection relationship for adjacent basic nodes according to the transmission sequence. The transmission connection relationship includes the transmission direction, the transmission triggering condition and the triggering interval threshold. The transmission triggering condition is that the actual state of the operation link unit or risk factor corresponding to the previous basic node meets the preset triggering threshold. The triggering interval threshold is the average value of the triggering interval time corresponding to adjacent basic nodes in the risk evolution path.

[0062] Based on the transmission order in the basic node trigger sequence table, a transmission connection relationship is established for adjacent basic nodes. Taking basic nodes CR-005 and CR-008 as an example, the transmission direction is determined to be CR-005→CR-008. The transmission trigger condition is set when the actual state of risk factor A "mistakenly operating a disconnector as a circuit breaker" corresponding to the previous basic node CR-005 meets the preset trigger threshold, such as when the operator fails to pass the device authentication of the five-prevention system once or more. The trigger interval threshold is calculated as follows: collect all risk evolution paths containing the transmission connection relationship between CR-005 and CR-008, extract the interval between the trigger time of CR-005 and the trigger time of CR-008 in each path, calculate the average of these intervals, and use this average as the trigger interval threshold. For example, if the average interval is 30 seconds, the trigger interval threshold is set to 30 seconds. For other adjacent basic nodes, a transmission connection relationship including transmission direction, transmission trigger condition, and trigger interval threshold is established in the same way.

[0063] Step S135: Structure and integrate all basic nodes and the transmission connection relationships between nodes, establish a node index table according to the node number, construct a path topology diagram according to the transmission direction, and form a switching operation risk association network containing the node hierarchy structure and transmission path topology.

[0064] All basic node information (operation step identifier, risk factor identifier, trigger direction identifier, association basis record, node number, etc.) and the transmission connection relationships between nodes (transmission direction, transmission trigger condition, trigger interval threshold, etc.) are structurally integrated. A node index table is established, arranged in ascending order of node number. Each entry in the index table corresponds to a basic node, containing the node number and all attribute information of that node, facilitating quick query of node details. A path topology graph is constructed in the form of a directed graph according to the transmission direction. Each node in the graph represents a basic node, the directed edges represent the transmission connection relationship, the arrow direction indicates the transmission direction, and the transmission trigger condition and trigger interval threshold are marked on the edges. By combining the node index table and the path topology graph, a switching operation risk association network is formed. This switching operation risk association network clearly shows the node hierarchy structure (such as starting node, intermediate node, and ending node) and the transmission path topology (the connection relationship and transmission direction between each node).

[0065] Step S136: Add attribute tags to each node in the switching operation risk association network. The attribute tags include node type identifier, initial value of association strength and status snapshot reference identifier. The node type identifier distinguishes between operation link nodes and risk factor nodes. The status snapshot reference identifier points to the corresponding equipment status snapshot and environmental status snapshot in the historical operation risk information. Verify the completeness of the attribute tag information of each node and correct any abnormal situations where information is missing or inconsistent.

[0066] Step S1361: Traverse all basic nodes in the risk association network of switching operation, read the operation link identifier, risk factor identifier and trigger direction identifier contained in each basic node, and identify the core information corresponding to each basic node. The core information is the operation link identifier or risk factor identifier.

[0067] Iterate through all basic nodes in the risk association network of switching operations. For each basic node, read its operation step identifier, risk factor identifier, and trigger direction identifier. If the trigger direction identifier is "C→R", it means that the core information of this basic node is the operation step identifier, because the operation step unit triggers the risk factor; if the trigger direction identifier is "R→C", the core information is the risk factor identifier, because the risk factor affects the operation step unit. For example, the trigger direction identifier of basic node CR-005 is "C→R", and its core information is the operation step identifier "operator did not strictly follow the operation ticket verification process"; the core information of basic node RC-010 (assuming the trigger direction identifier is "R→C") is the risk factor identifier "ambient temperature is too high".

[0068] Step S1362: Assign a node type identifier to each basic node according to the type of core information. When the core information is an operation step identifier, the node type identifier is set as the operation step node identifier, and the operation action description corresponding to the operation step identifier is recorded. When the core information is a risk factor identifier, the node type identifier is set as the risk factor node identifier, and the risk type description corresponding to the risk factor identifier is recorded.

[0069] For basic nodes whose core information is an operation step identifier, the node type identifier is set to the prefix "OP-" followed by the node number. For example, the node type identifier for basic node CR-005 is "OP-CR-005". Simultaneously, the description of the operation action corresponding to the operation step identifier is recorded, namely, "The operator did not strictly follow the operation ticket verification process and did not verify the equipment number and status." For basic nodes whose core information is a risk factor identifier, the node type identifier is set to the prefix "RF-" followed by the node number. For example, the node type identifier for basic node RC-010 is "RF-RC-010". Simultaneously, the description of the risk type corresponding to the risk factor identifier is recorded, namely, "The ambient temperature exceeds the normal operating temperature range of the equipment, which may lead to equipment insulation aging and personnel heatstroke."

[0070] Step S1363: Extract the associated records corresponding to each basic node from the historical operational risk information. The associated records are the number of times the basic node and other nodes appear together in historical risk events. The associated records are weighted according to the severity of the historical risk events. The higher the severity of the historical risk event, the greater the weight of the corresponding associated record.

[0071] The system retrieves the occurrence of each basic node in historical risk events from historical operational risk information, and counts the number of times each basic node co-occurred with other nodes as a correlation record. For example, basic node CR-005 participated in 15 historical risk events in the past five years, co-occurring with basic node CR-008 8 times and with basic node CR-015 5 times, etc. Weights are assigned according to the severity of historical risk events, categorized into four levels: extremely serious, serious, relatively serious, and moderate, with corresponding weight values ​​of 4, 3, 2, and 1, respectively. If a historical risk event is classified as serious, and basic nodes CR-005 and CR-008 co-occurred in that event, the weighted value of this correlation record is 3. The weighted values ​​of all correlation records are summed to obtain the weighted statistical result.

[0072] Step S1364: Calculate the ratio of the weighted associated record count of each basic node to the average weighted associated record count of all nodes to obtain the association frequency coefficient. At the same time, calculate the probability of occurrence of the basic node in all historical risk events. The probability of occurrence is the ratio of the number of historical risk events that the basic node has occurred to the total number of historical risk events.

[0073] Assuming there are 100 basic nodes in the risk association network for switching operations, and the total number of weighted association records for all nodes is 5000, then the average number of weighted association records is 50. If the weighted association record count for basic node CR-005 is 80, then its association frequency coefficient is 80 divided by 50, resulting in 1.6. If the total number of historical risk events is 200, and basic node CR-005 has participated in 15 historical risk events, then its probability of occurrence is 15 divided by 200, resulting in 0.075.

[0074] Step S1365: Determine the initial value of the association strength of each basic node based on the association frequency coefficient and the occurrence probability. The higher the association frequency coefficient and the greater the occurrence probability, the greater the initial value of the association strength. Establish a correspondence table between the association frequency coefficient, the occurrence probability and the initial value of the association strength.

[0075] The initial value of the association strength is calculated by multiplying the association frequency coefficient by the probability of occurrence by a constant (this constant is adjusted according to the actual application scenario, and the specific value is not involved here). For the basic node CR-005, the association frequency coefficient is 1.6, the probability of occurrence is 0.075, and its initial value of association strength is 1.6 multiplied by 0.075 and then multiplied by the constant. The calculation results of the initial values ​​of association strength corresponding to different combinations of association frequency coefficients and probabilities of occurrence are compiled to establish a correspondence table between association frequency coefficients, probabilities of occurrence, and initial values ​​of association strength. For example, the initial value of association strength corresponding to association frequency coefficients of 1.5-2.0 and probabilities of occurrence of 0.05-0.10 is 0.15-0.25 (the specific value is obtained through the above calculation rule).

[0076] Step S1366: Integrate the node type identifier, initial value of association strength, core information description, association frequency coefficient and occurrence probability into a unified format attribute label. The attribute label is bound to the corresponding basic node through the node number. Through the uniqueness verification of the node number, each node number uniquely corresponds to one attribute label.

[0077] The node type identifier "OP-CR-005" of the basic node CR-005, the initial association strength value (e.g., 0.20), the core information description "Operator did not strictly follow the operation ticket verification process, and did not verify the equipment number and status", the association frequency coefficient 1.6, and the occurrence probability 0.075 are integrated into a single attribute tag. The attribute tag is in key-value pair format, such as "Node type identifier: OP-CR-005; Initial association strength value: 0.20; Core information description: Operator did not strictly follow the operation ticket verification process, and did not verify the equipment number and status; Association frequency coefficient: 1.6; Occurrence probability: 0.075". This attribute tag is bound to the basic node through the node number "CR-005" to ensure that each node number uniquely corresponds to one attribute tag and avoid confusion.

[0078] Step S1367: Perform attribute validation on the bound basic nodes, check whether the node type identifier in the attribute tags is consistent with the core information, whether the calculation process of the initial value of the association strength is complete, and whether the association frequency coefficient and the probability of occurrence are reasonable. Correct any abnormalities such as mismatch between the attribute tags and the node core information, missing calculation processes, or unreasonable values, and form the final set of node attribute tags.

[0079] Validate the basic nodes after binding attribute tags. For basic node CR-005, check whether the node type identifier "OP-CR-005" (operation step node identifier) ​​is consistent with the core information "Operator did not strictly follow the operation ticket verification process" (operation step identifier). If they are consistent, the validation passes. Check whether the calculation process of the initial value of association strength 0.20 is complete and whether it can be calculated from the association frequency coefficient 1.6, the probability of occurrence 0.075, and the preset constant. If the calculation process is complete, the validation passes. Analyze whether the association frequency coefficient 1.6 and the probability of occurrence 0.075 are within a reasonable range and compare them with the values ​​of other similar basic nodes. If they are within the normal distribution range, they are considered reasonable. If it is found that the node type identifier of a basic node is a risk factor node identifier in the attribute tag, but the core information is an operation step identifier, then correct the node type identifier; if the calculation process of the initial value of association strength is missing, then supplement the calculation process; if the association frequency coefficient or the probability of occurrence is obviously abnormal (such as exceeding the normal range by several times), then re-check the historical data statistics process and correct the abnormal values. After the above verification and correction, the final set of node attribute labels is formed.

[0080] Step S140: Perform risk transmission intensity analysis on the nodes and transmission relationships in the switching operation risk association network, and generate risk transmission weight descriptions for each associated link.

[0081] Step S141: Extract the actual occurrence frequency data corresponding to each node in the risk association network of switching operation from the historical operation risk information, perform hierarchical statistics on the occurrence frequency data, and form a multi-time period frequency distribution table. The occurrence frequency data is the number of times the operation link unit or risk factor corresponding to the node appears in the historical operation.

[0082] For each node in the switching operation risk association network, such as node CR-005 (operation unit "operator did not strictly follow the operation ticket verification process") and node RF-RC-010 (risk factor "excessively high ambient temperature"), the corresponding actual occurrence frequency data is extracted from historical operation risk information. Actual occurrence frequency data refers to the number of times the corresponding operation unit is executed or the number of times the risk factor occurs. The occurrence frequency data is stratified by time, such as by year, quarter, or month, forming a multi-period frequency distribution table. Taking node CR-005 as an example, the multi-period frequency distribution table records that it occurred 3 times in 2019, 5 times in 2020, 2 times in 2021, 1 time in the first quarter of 2022, and 2 times in the second quarter, clearly showing the frequency changes of this node in different time periods.

[0083] Step S142: Analyze the correlation between the occurrence frequency data of each node and the number of historical risk events. Calculate the ratio of the number of risk events occurring within a preset time window after the node appears to the total number of times the node appears, as the correlation degree. At the same time, analyze the time change trend of the correlation degree in conjunction with the frequency distribution table of multiple time periods.

[0084] For node CR-005, the preset time window is set to one hour after its occurrence. The total number of occurrences of node CR-005 is 50, of which 15 occurred within one hour of its occurrence. Therefore, the correlation degree is 15 divided by 50, resulting in 0.3. The temporal trend of the correlation degree is analyzed using a multi-period frequency distribution table. The correlation degree for each year is calculated, for example, 0.25 in 2019, 0.35 in 2020, and 0.28 in 2021. The changing patterns of the correlation degree values ​​in these years are observed to determine whether they show an upward trend, a downward trend, or a fluctuating trend. The same method is used to calculate the correlation degree and analyze its temporal trend for node RF-RC-010.

[0085] Step S143: Assign a node importance coefficient to each node based on the degree of correlation and the time trend of the degree of correlation. Record the calculation basis and time trend analysis results of the node importance coefficient. The higher the degree of correlation and the more stable the time trend, the greater the node importance coefficient.

[0086] The node importance coefficient is set to a range of 0 to 1. For node CR-005, with a correlation level of 0.3 and a relatively stable time trend (the annual correlation level fluctuation is within ±0.05), a higher node importance coefficient is assigned, such as 0.6. For nodes with a correlation level of 0.2 and a large fluctuation in the time trend (the annual correlation level fluctuation exceeds ±0.1), a lower node importance coefficient is assigned, such as 0.3. The basis for calculating the node importance coefficient is recorded, such as the basis for calculating node CR-005 being a correlation level of 0.3 and a stable time trend; the time trend analysis results are also recorded, such as "the correlation level fluctuated little from 2019 to 2021, showing a slight upward trend overall."

[0087] Step S144: Analyze the historical trigger success rate corresponding to each transmission connection in the switching operation risk association network, query all trigger records corresponding to the transmission connection in the historical operation risk information. The trigger record includes the trigger condition satisfaction status, the actual trigger status of the next node, and the parameter change data at the time of triggering. Count the number of valid records where the trigger condition is satisfied and the next node is successfully triggered, as well as the total number of records where the trigger condition is satisfied. Calculate the ratio of the number of valid records to the total number of records as the trigger success rate.

[0088] Taking the conductive connection CR-005→CR-008 as an example, query all trigger records involving this conductive connection in the historical operational risk information. Each trigger record includes the fulfillment of trigger conditions (e.g., whether the operator failed the five-prevention system verification once), the actual triggering status of the subsequent node CR-008 (whether "disconnector operation under load" occurred), and parameter changes during triggering (e.g., load current value during operation, alarm information from the five-prevention system, etc.). The total number of records where the trigger conditions were met was 100, of which 60 were valid records of successful triggering of the subsequent node CR-008. Therefore, the trigger success rate for this conductive connection is 60 divided by 100, resulting in 0.6. The trigger success rate for all conductive connections in the network is calculated using this method.

[0089] Step S145: Combine the node importance coefficient and the trigger success rate of the transmission connection relationship to calculate the transmission strength value of each associated link. The transmission strength value is the product of the node importance coefficient of the previous node and the trigger success rate of the corresponding transmission connection relationship. At the same time, refer to the ratio of the parameter change data amplitude in the trigger record to the historical average parameter change amplitude to fine-tune the transmission strength value. The larger the ratio of the parameter change data amplitude, the larger the fine-tuning amplitude.

[0090] Step S1451: Extract the node numbers of the preceding and following nodes corresponding to each associated link, query the node attribute tags in the switching operation risk association network through the node numbers, and obtain the node importance coefficient and calculation basis of the preceding node.

[0091] For the related link CR-005→CR-008, the preceding node is CR-005 and the following node is CR-008. By querying the node attribute label through the node number CR-005, its node importance coefficient is found to be 0.6, calculated based on the degree of association of 0.3 and a stable time change trend.

[0092] Step S1452: Query all trigger records of the transmission connection relationship corresponding to the associated link in the historical operation risk information, sort the trigger records in order of trigger time, and each trigger record includes the trigger condition satisfaction status, the actual trigger status of the next node, the equipment parameter data at the time of triggering, the environmental parameter data, and the description of the trigger result.

[0093] A total of 100 trigger records were retrieved for the conductive connection CR-005→CR-008, sorted from earliest to most recent in chronological order of trigger time. Each trigger record details whether the trigger conditions were met (e.g., "yes" or "no"), the actual trigger status of the subsequent node ("successfully triggered" or "not triggered"), the equipment parameter data at the time of triggering (e.g., voltage and current values ​​on both sides of the disconnector switch), the environmental parameter data (e.g., ambient temperature and humidity at the time), and a description of the trigger result (e.g., "caused an electric arc" or "did not cause any abnormal consequences").

[0094] Step S1453: Count the number of valid records in the trigger record where the trigger condition is met and the next node is successfully triggered, mark the normal range deviation values ​​of the device parameter data and environmental parameter data in the valid records, and at the same time count the total number of records that meet the trigger condition, and distinguish the number of records in the total number of records that fail to trigger due to excessive parameter deviation.

[0095] Of the 100 trigger records, 80 records met the trigger conditions (the total number of records in step S144 is corrected here to maintain logical consistency), of which 60 are valid records where the subsequent node was successfully triggered. For these 60 valid records, the deviation of the equipment parameter data (such as current value) from its normal operating range is marked. For example, if the normal current range is 0-5A and the current value in a valid record is 10A, then the deviation is 5A; if the normal range of environmental parameter data (such as temperature) is -5℃-40℃ and the temperature in a record is 45℃, then the deviation is 5℃. Of the 80 records where the trigger conditions were met, 20 records did not trigger the subsequent node. Further analysis revealed that 10 records failed to trigger due to excessive deviation of equipment parameters (such as current exceeding the tripping threshold set by the five-prevention system, automatically preventing operation).

[0096] Step S1454: Calculate the ratio of the number of valid records to the total number of records to obtain the triggering success rate of the conductive connection relationship. At the same time, calculate the average value of the normal range deviation of the device parameter data and environmental parameter data in the valid records as the parameter deviation reference value, and calculate the average value of the parameter deviation in all historical valid records as the historical average parameter deviation.

[0097] The number of valid records is 60, and the total number of records meeting the trigger condition is 80. The trigger success rate is 60 divided by 80, which equals 0.75. Calculate the average of the normal range deviation values ​​of the equipment parameter data from the 60 valid records. For example, the average current deviation is 3A, and the average temperature deviation is 4℃. Combine these averages as the parameter deviation reference value (a weighted average method can be used, with weights set according to parameter importance). Calculate the average of the normal range deviation values ​​of the equipment parameter data and environmental parameter data from all historical valid records (not limited to this conductive connection relationship) as the historical average parameter deviation. For example, the historical average current deviation is 2A, and the historical average temperature deviation is 3℃.

[0098] Step S1455: Multiply the node importance coefficient of the previous node with the trigger success rate to obtain the initial conduction strength value. Then, fine-tune the initial conduction strength value according to the ratio of the parameter deviation reference value to the historical average parameter deviation. The smaller the parameter deviation reference value ratio, the larger the conduction strength value after fine-tuning; the larger the parameter deviation reference value ratio, the smaller the conduction strength value after fine-tuning.

[0099] The previous node, CR-005, has a node importance coefficient of 0.6 and a trigger success rate of 0.75. The initial conduction strength value is 0.6 multiplied by 0.75, resulting in 0.45. The ratio of the parameter deviation reference value to the historical average parameter deviation is calculated. For example, the ratio of the current deviation reference value of 3A to the historical average current deviation of 2A is 1.5, and the ratio of the temperature deviation reference value of 4℃ to the historical average temperature deviation of 3℃ is 1.33. The average of the comprehensive ratios is 1.415. The fine-tuning rule is set as follows: when the ratio is greater than 1, the fine-tuning coefficient is (2 - ratio); when the ratio is less than or equal to 1, the fine-tuning coefficient is (1 + (1 - ratio)). Here, the ratio 1.415 is greater than 1, so the fine-tuning coefficient is 2 - 1.415 = 0.585. The fine-tuned conduction strength value is the initial conduction strength value multiplied by the fine-tuning coefficient, i.e., 0.45 multiplied by 0.585, yielding the fine-tuned conduction strength value (the specific value is calculated and not shown here).

[0100] Step S1456: Perform range verification on the fine-tuned conduction intensity value, compare it with the preset reasonable conduction intensity range, eliminate abnormal values ​​that exceed the reasonable range, analyze the cause of abnormal values, if it is caused by abnormal trigger record data, re-count the trigger records, if it is caused by calculation logic deviation, correct the calculation logic, and retain the conduction intensity value that conforms to the actual risk conduction logic.

[0101] The preset reasonable transmission intensity range is 0 to 1. The fine-tuned transmission intensity value is compared with this range. If the value is between 0 and 1, it passes the range check. If the value is less than 0 or greater than 1, it is considered an abnormal value. The cause of the abnormal value is analyzed. If it is found that some trigger record data was entered incorrectly (e.g., a failed trigger record was mistakenly recorded as a successful trigger), the trigger record data is re-counted, and the transmission intensity value is recalculated. If it is found that there is a flaw in the fine-tuning coefficient calculation logic (e.g., the correspondence between the ratio and the fine-tuning coefficient is unreasonable), the calculation logic is corrected, and the fine-tuning calculation is performed again. After verification and correction, the transmission intensity value that conforms to the actual risk transmission logic is retained.

[0102] Step S1457: Use the retained transmission strength value as the final transmission strength value for each associated link, record the calculation process of the final transmission strength value, the parameter deviation reference value ratio and the range verification result. The final transmission strength value directly reflects the influence of the associated link in the risk transmission process.

[0103] The transmission strength value that passes the range verification is determined as the final transmission strength value of the associated link. The calculation process of this value is recorded in detail, including obtaining the node importance coefficient, calculating the trigger success rate, calculating the parameter deviation reference value ratio, and fine-tuning the process; the specific value of the parameter deviation reference value ratio and the range verification result (e.g., "Verification passed, value is within a reasonable range") are also recorded. A higher final transmission strength value indicates a greater influence of the associated link in the process of risk transmission from one node to another, meaning it is more likely to trigger risk events in subsequent nodes.

[0104] Step S146: Normalize the transmission intensity values, divide the weight levels according to the normalization results, and generate a risk transmission weight description for each associated link. The risk transmission weight description includes the weight value, weight level identifier, calculation basis, and fine-tuning instructions.

[0105] The conduction strength values ​​of all associated links are normalized, mapping the values ​​to the range of 0 to 1. Linear normalization is used, meaning the normalized value for each associated link is equal to (the conduction strength value of that link - the minimum conduction strength value of all links) divided by (the maximum conduction strength value of all links - the minimum conduction strength value of all links). Weight levels are assigned based on the normalization results. For example, normalized values ​​of 0-0.2 are low weight (level L), 0.2-0.5 are medium weight (level M), 0.5-0.8 are high weight (level H), and 0.8-1 are high weight (level VH). For the associated link CR-005→CR-008, the normalized weight value is 0.6, placing it in the high weight level (H). The risk transmission weight description includes a weight value of 0.6, a weight level identifier H, a calculation basis (node ​​importance coefficient 0.6, trigger success rate 0.75, parameter deviation reference value ratio 1.415, etc.) and fine-tuning instructions (a fine-tuning coefficient of 0.585 is calculated based on the parameter deviation ratio to fine-tune the initial transmission strength value).

[0106] Step S150: Combining the risk transmission weight description with the real-time execution information of the current switching operation, the risk evolution path in the switching operation risk association network is deduced, and the risk prediction results of the substation switching operation are output.

[0107] Step S151: Analyze the real-time execution information of the current switching operation, extract the currently executing operation unit, the corresponding equipment operating status information, environmental status information and execution progress data, and add real-time collection timestamps to the equipment operating status information and environmental status information.

[0108] The current switching operation is a power outage operation for this 10kV line bay. Real-time execution information is obtained through the real-time data transmission module of the substation monitoring system. The real-time execution information is parsed, and the currently executing operation unit is extracted as "opening the 10kV line disconnect switch" (time sequence identifier "002"). The corresponding equipment operating status information includes the current position signal of the disconnect switch (open / closed), auxiliary contact status, and motor current of the operating mechanism; environmental status information includes the ambient temperature, relative humidity, and wind speed at the current operating site; and execution progress data includes the planned execution time, already executed time, and remaining time for this operation unit. Real-time acquisition timestamps are added to the equipment operating status information and environmental status information, accurate to the millisecond level, to ensure the time accuracy of the data.

[0109] Step S152: Match the identifier of the currently executing operation unit with the operation identifier of the operation node in the switching operation risk association network to determine the corresponding target node, and obtain the attribute label, initial value of association strength and association basis record of the target node.

[0110] The currently executing operation unit is identified as "002 - Opening the 10kV line disconnect switch". Matching is performed within the operation unit nodes of the switching operation risk association network. The corresponding target node is found; assuming node number CR-008, its operation unit identifier matches the current operation unit identifier. The attribute tags of this target node are obtained, including the node type identifier "OP-CR-008", the initial association strength value of 0.5, and the core information description "Opening the 10kV line disconnect switch operation", etc.; the initial association strength value is 0.5; the association basis records include historical records of abnormal equipment parameters and environmental change data related to risk events occurring in this operation unit.

[0111] Step S153: Based on the risk transmission weight description, starting from the target node, according to the transmission path topology in the switching operation risk association network, traverse all transmission paths starting from the target node, and record the node sequence, transmission connection relationship and corresponding risk transmission weight description of each transmission path.

[0112] Starting from the target node CR-008, all transmission paths originating from CR-008 are traversed according to the transmission path topology of the switching operation risk association network. For example, the first transmission path is CR-008→CR-012→CR-018→RF-025; the second transmission path is CR-008→CR-020→RF-030; the third transmission path is CR-008→RF-022, and so on. For each transmission path, the sequence of nodes it contains (such as CR-008, CR-012, CR-018, RF-025), the transmission connection relationship between each node (such as the transmission direction, triggering condition, and triggering interval threshold of CR-008→CR-012), and the corresponding risk transmission weight description (such as the weight value of CR-008→CR-012 being 0.5 and the level identifier M, and the weight value of CR-012→CR-018 being 0.7 and the level identifier H, etc.).

[0113] Step S154: Based on the current device operating status information, environmental status information, and real-time collection timestamp, compare the historical status data corresponding to the status snapshot reference identifier in the attribute label of the subsequent nodes on each transmission path, determine whether the triggering conditions of the subsequent nodes on each transmission path are met, and record the subsequent nodes that meet the triggering conditions and the corresponding judgment basis.

[0114] Step S1541: Extract the conduction trigger condition description corresponding to the subsequent nodes on each conduction path. The conduction trigger condition description includes the state range of equipment operating parameters, the state range of environmental parameters, the standard requirements for operation execution, and the allowable range of parameter deviation.

[0115] Taking the conduction path CR-008→CR-012→CR-018→RF-025 as an example, subsequent nodes include CR-012, CR-018, and RF-025. The description of the conduction triggering conditions for CR-012 is extracted as follows: the equipment operating parameter range is that the load current during disconnection switch operation is greater than 5A; the environmental parameter range is that the ambient temperature is greater than 35℃; the operation execution specification requires that the operator is not wearing insulated gloves; and the allowable parameter deviation range is that the load current deviation does not exceed ±1A and the temperature deviation does not exceed ±2℃.

[0116] Step S1542: Extract the actual parameter data corresponding to the transmission trigger condition description from the real-time execution information of the current switching operation. The actual parameter data includes the actual values ​​of the current equipment operating parameters, the actual values ​​of the environmental parameters, the actual status of the operation execution, and the parameter acquisition time. The parameter acquisition time is consistent with the real-time acquisition timestamp.

[0117] Extract the actual parameter data corresponding to the CR-012 conduction trigger condition from the current real-time execution information: the actual value of the disconnect switch load current in the equipment operation parameters is obtained through the measurement value of the secondary side of the current transformer; the actual value of the ambient temperature in the environmental parameters is obtained through the on-site temperature and humidity sensor; the actual status of the operation execution is confirmed by monitoring video and the operation record of the five-prevention system to confirm whether the operator is wearing insulating gloves; the parameter acquisition time and the real-time acquisition timestamp are both the current operation time to ensure time consistency.

[0118] Step S1543: Compare the actual parameter data with the state range or specification requirements in the description of the conduction triggering conditions one by one. Compare the actual values ​​of the equipment operating parameters with the actual values ​​of the environmental parameters to the corresponding state range. Compare the actual state of the operation execution with the specification requirements of the operation execution. Record the deviation of each actual parameter data from the corresponding requirements.

[0119] Compare the actual load current of the disconnecting switch with the state range "greater than 5A". If the actual value is 8A, the state range requirement is met, and the deviation is "positive deviation 3A". Compare the actual ambient temperature with "greater than 35℃". If the actual temperature is 38℃, the requirement is met, and the deviation is "positive deviation 3℃". Compare the actual operation status "operator not wearing insulating gloves" with the standard requirement "not wearing insulating gloves" (the standard requirement here may be "insulating gloves should be worn", in which case the actual status does not meet the requirement and needs to be determined according to the actual standard). Record the deviation as "does not meet the standard requirements" or "meets the standard requirements".

[0120] Step S1544: Count the number of parameter items that meet the requirements in the transmission trigger conditions of subsequent nodes on each transmission path. The parameter items that meet the requirements are those whose actual parameter data is within the state range or specification requirements. At the same time, count the total number of parameter items included in the transmission trigger conditions and distinguish between critical parameter items and non-critical parameter items in the total parameter items.

[0121] Assume the conduction triggering conditions for CR-012 include three parameters: load current, ambient temperature, and operating procedures (wearing insulated gloves), for a total of three parameters. Load current and operating procedures are critical parameters, while ambient temperature is a non-critical parameter. If the actual parameter data shows that the load current and ambient temperature meet the requirements, but the operating procedures do not, then the number of parameters that meet the requirements is two.

[0122] Step S1545: Calculate the ratio of the number of qualified parameter items to the total number of parameter items to obtain the trigger condition satisfaction. If there are key parameter items, calculate the ratio of the number of qualified key parameter items to the total number of key parameter items separately as the key parameter satisfaction.

[0123] The trigger condition satisfaction rate is calculated by dividing the number of compliant parameter items (2) by the total number of parameter items (3), resulting in 0.67. There are 2 critical parameter items, and 1 of them meets the requirements (load current meets the requirements, but operating specifications do not). The critical parameter satisfaction rate is calculated by dividing 1 by 2, resulting in 0.5.

[0124] Step S1546: Set the trigger condition satisfaction threshold and the key parameter satisfaction threshold. When the trigger condition satisfaction reaches or exceeds the trigger condition satisfaction threshold and the key parameter satisfaction reaches or exceeds the key parameter satisfaction threshold, determine that the trigger condition of the subsequent node on the transmission path is satisfied.

[0125] The threshold for trigger condition satisfaction is set to 0.6, and the threshold for key parameter satisfaction is set to 0.5. For CR-012, the trigger condition satisfaction of 0.67 reaches the threshold of 0.6, and the key parameter satisfaction of 0.5 reaches the threshold of 0.5. Therefore, it is determined that the trigger condition of this subsequent node is satisfied.

[0126] Step S1547: When the trigger condition satisfaction is lower than the trigger condition satisfaction threshold or the key parameter satisfaction is lower than the key parameter satisfaction threshold, it is determined that the trigger condition of the subsequent node on the transmission path is not met. Record the unmet parameter items, parameter item types, deviation values ​​between actual parameter data and requirements, and analysis of the reasons for deviation. Parameter item types are divided into key parameter items and non-key parameter items.

[0127] If the trigger condition satisfaction rate of another subsequent node CR-020 is 0.5 (below the threshold of 0.6), and the critical parameter satisfaction rate is 0.4 (below the threshold of 0.5), then its trigger condition is determined to be unmet. Record the unmet parameter items, such as "low operating mechanism pressure" (critical parameter item) and "high ambient humidity" (non-critical parameter item); the parameter item type is marked as critical and non-critical respectively; the deviation value between the actual parameter data and the requirement, such as the actual operating mechanism pressure being 0.5MPa lower than the required value, and the actual ambient humidity being 10% higher than the required value; the analysis of the deviation reasons, such as "the operating mechanism may have air leakage leading to low pressure" or "today's rainfall has led to increased ambient humidity".

[0128] Step S155: Assess the risk level of the transmission path that meets the triggering conditions. The risk level assessment is calculated by accumulating the weight values ​​in the risk transmission weight description of each associated link on the path, while also referring to the initial value of the association strength of the nodes in the transmission path. The larger the accumulated result, the higher the risk level.

[0129] For a transmission path where the triggering condition is met, such as CR-008→CR-012→CR-018→RF-025, the weight values ​​in the risk transmission weight descriptions of each associated link on the path are extracted: the weight value for CR-008→CR-012 is 0.7, the weight value for CR-012→CR-018 is 0.6, and the weight value for CR-018→RF-025 is 0.8. The above weight values ​​are then summed, resulting in a total of 0.7 + 0.6 + 0.8 = 2.1. Simultaneously, referring to the initial values ​​of the association strength of each node in this transmission path, such as 0.5 for CR-008, 0.4 for CR-012, 0.3 for CR-018, and 0.6 for RF-025, the initial values ​​of the association strength are also accumulated (or combined with the weight values ​​in a certain proportion; a simple accumulation example is used here), resulting in an accumulated association strength of 0.5 + 0.4 + 0.3 + 0.6 = 1.8. Combining the accumulated weight value and the accumulated association strength value (e.g., adding them together yields 3.9), the larger the accumulated value, the higher the corresponding risk level.

[0130] Step S156: Select the transmission path where the risk level assessment result exceeds the preset risk threshold as the risk evolution path, and mark the key nodes and triggering conditions in the risk evolution path. The key node is the node corresponding to the associated link with the highest weight value.

[0131] The preset risk threshold is set according to the safety management requirements of the substation, such as a comprehensive cumulative threshold of 3.0. The comprehensive cumulative result of the transmission path CR-008→CR-012→CR-018→RF-025 is 3.9, which exceeds the preset risk threshold of 3.0, and therefore it is selected as a risk evolution path. In this risk evolution path, the weight values ​​of each associated link are 0.7, 0.6, and 0.8, respectively. The associated link with the highest weight value is CR-018→RF-025 (0.8). The nodes corresponding to this associated link are CR-018 and RF-025, and these two nodes are marked as critical nodes. At the same time, the triggering conditions of the critical nodes are marked, such as the triggering condition of CR-018 being "arc duration exceeds 2 seconds", and the triggering condition of RF-025 being "phase-to-phase short-circuit current reaches the protection action value".

[0132] Step S157: Integrate the risk factors, potential risk event types, key node information, and triggering conditions corresponding to all risk evolution paths to generate risk prediction results for substation switching operations that include a risk occurrence probability description, a risk impact range description, risk triggering node information, and risk prevention and control recommendations. The risk occurrence probability description is determined based on the ratio of the number of risk evolution paths to the total number of transmission paths, and the risk impact range description is determined based on the equipment range and operation range involved in the nodes of the risk evolution path.

[0133] The system integrates information on all selected risk evolution paths. For example, besides the paths mentioned above, there is another risk evolution path, CR-008→RF-022, whose corresponding risk factor is "overvoltage during disconnector operation," and the potential risk event type is "equipment insulation breakdown." Key node information includes the key node number, node type, and initial correlation strength value for each path; triggering conditions include the specific triggering parameter requirements for each key node. The probability of risk occurrence is described by the ratio (2 / 5) of the number of risk evolution paths (e.g., 2) to the total number of conduction paths (e.g., 5). The scope of risk impact is described based on the equipment range (e.g., 10kV line disconnectors, circuit breakers, busbars) and operational range (power outage operation of the 10kV line bay) involved in the nodes of the risk evolution path. Risk prevention and control recommendations are proposed for each risk evolution path. For example, for the path CR-008→CR-012→CR-018→RF-025, it is recommended to immediately stop operation, check the operating mechanism of the disconnecting switch and the load condition, and ensure that operators wear appropriate insulated protective equipment. For the path CR-008→RF-022, it is recommended to take overvoltage protection measures, such as activating surge arresters. After integrating the above information, a complete risk prediction result for substation switching operations is generated.

[0134] Based on the same inventive concept, please refer to Figure 2 This paper shows a schematic block diagram of a substation switching operation risk prediction system 100 based on big data analysis, which is provided in an embodiment of this application for executing the above-described substation switching operation risk prediction method based on big data analysis. The substation switching operation risk prediction system 100 based on big data analysis may include a communication unit 110, a machine-readable storage medium 120, and a processor 130.

[0135] In this embodiment, both the machine-readable storage medium 120 and the processor 130 are located in the substation switching operation risk prediction system 100 based on big data analysis and are separately configured. Alternatively, the machine-readable storage medium 120 can also be integrated into the processor 130 and can communicate and interact with external systems through the communication unit 110. The machine-readable storage medium 120 is used to store machine-executable instructions for executing the scheme of this application, and the processor 130 is used to execute the machine-executable instructions stored in the machine-readable storage medium 120 to implement the substation switching operation risk prediction method based on big data analysis provided in the aforementioned method embodiment.

[0136] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A substation switching operation risk prediction method based on big data analysis, characterized in that, The method comprises: Collecting multi-source associated information related to substation switching operation, wherein the multi-source associated information comprises operation execution information, device operation information, environmental influence information and historical operation risk information; Constructing risk factor associated links for the multi-source associated information to generate a set of one-way associated links between the operation and risk factors, wherein each associated link in the set of one-way associated links corresponds to a mapping relationship between a single operation link and a single risk factor; Based on the set of one-way associated links and the risk evolution path in the historical operation risk information, constructing a switching operation risk associated network, wherein the switching operation risk associated network comprises associated link nodes and conduction relationship descriptions between the nodes; Analyzing the risk conduction strength of the nodes and conduction relationship in the switching operation risk associated network to generate risk conduction weight descriptions corresponding to each associated link; Combining the risk conduction weight descriptions and real-time execution information of the current switching operation, deducing the risk evolution path in the switching operation risk associated network, and outputting the risk prediction result of the substation switching operation.

2. The substation switching operation risk prediction method based on big data analysis according to claim 1, characterized in that, The method of constructing risk factor associated links for the multi-source associated information to generate a set of one-way associated links between the operation and risk factors comprises: Splitting the operation execution information in the multi-source associated information into a plurality of continuous operation link units, each operation link unit corresponding to a specific execution action in the switching operation, and adding a time sequence identifier and a device association identifier to each operation link unit; From the device operation information, environmental influence information and historical operation risk information in the multi-source associated information, screening the risk factors associated with each operation link unit, first matching the abnormal parameter records in the device operation information through the device association identifier, then matching the same period environmental change data in the environmental influence information through the time sequence identifier, and finally matching the same type operation risk records in the historical operation risk information through the action type of the operation link unit, forming a risk factor list corresponding to each operation link unit; Analyzing the triggering relationship of each operation link unit and each risk factor in the corresponding risk factor list, determining the direction attribute of the triggering relationship by comparing the order of the operation link unit execution time and the risk factor occurrence time in the historical operation risk information, wherein the direction attribute is that the operation link unit triggers the risk factor or the risk factor affects the operation link unit; According to the direction attribute of the triggering relationship, constructing a one-way associated link between each operation link unit and the corresponding risk factor, wherein the one-way associated link comprises operation link identifier, risk factor identifier, triggering direction identifier and associated basis record, and the associated basis record is the matched device abnormal parameter record, environmental change data or historical risk record; Performing repeatability check on all one-way associated links, comparing the consistency of the combination of operation link identifier, risk factor identifier and triggering direction identifier, eliminating the one-way associated links with the same combination, and retaining the unique one-way associated links. The reserved unique one-way association link is sorted according to the time sequence of the operation link unit, and the corresponding execution order is marked, and the association basis record source of each one-way association link is marked, to form a one-way association link set between the operation and the risk factor. 3.The substation switching operation risk prediction method based on big data analysis of claim 1, wherein, The one-way association link set and the risk evolution path in the historical operation risk information are used to construct a switching operation risk association network, including: The risk evolution path corresponding to the occurred risk event is extracted from the historical operation risk information, each risk evolution path contains a plurality of risk factors and corresponding operation link unit sequences triggered in sequence, and the device state snapshot and the environment state snapshot when each risk factor and the operation link unit are triggered are recorded; Each one-way association link in the one-way association link set is taken as a basic node of the switching operation risk association network, and a unique node number is assigned to the basic node, the basic node includes an operation link identification, a risk factor identification, a trigger direction identification and an association basis record; The connection relationship between the adjacent basic nodes is established according to the conduction order, and the conduction connection relationship includes a conduction direction, a conduction trigger condition and a trigger interval threshold, the conduction trigger condition is that the actual state of the operation link unit or the risk factor corresponding to the previous basic node meets the preset trigger threshold, and the trigger interval threshold is the average value of the trigger interval time of the adjacent basic nodes in the risk evolution path; All the basic nodes and the conduction connection relationship between the nodes are structured and integrated, a node index table is established according to the node number, a path topology graph is constructed according to the conduction direction, and a switching operation risk association network including a node hierarchical structure and a conduction path topology is formed; Each node in the switching operation risk association network is added with an attribute label, the attribute label includes a node type identification, an association strength initial value and a state snapshot reference identification, the node type identification distinguishes the operation link node and the risk factor node, the state snapshot reference identification points to the corresponding device state snapshot and environment state snapshot in the historical operation risk information, the attribute label information integrity of each node is verified, and the abnormal situation of missing or inconsistent information is corrected. The connection relationship between the adjacent basic nodes is established according to the conduction order, and the conduction connection relationship includes a conduction direction, a conduction trigger condition and a trigger interval threshold, the conduction trigger condition is that the actual state of the operation link unit or the risk factor corresponding to the previous basic node meets the preset trigger threshold, and the trigger interval threshold is the average value of the trigger interval time of the adjacent basic nodes in the risk evolution path; 4. The substation switching operation risk prediction method based on big data analysis according to claim 3, characterized in that, Each risk evolution path is segmented, the risk evolution path is divided into a plurality of continuous trigger segments in chronological order, each trigger segment contains an operation link unit, a directly triggered risk factor and trigger time stamps corresponding to the operation link unit and the risk factor; Each trigger segment is matched with the one-way association link in the one-way association link set, the operation link unit and the risk factor combination in the trigger segment are compared, the basic node corresponding to each trigger segment is determined, and the node number of the matched basic node is recorded; ​ Corresponding to the adjacent trigger fragments in the same risk evolution path, the basis nodes are tracked, the associated basis records of the basis nodes corresponding to the previous trigger fragment and the basis nodes corresponding to the next trigger fragment are extracted, the correlation of the two in the device state parameters and the environment state parameters is analyzed, and the connection relationship between the output end of the previous basis node and the input end of the next basis node is determined; According to the connection relationship, the conduction direction between the basis nodes is determined, and the conduction direction is from the basis node triggered first to the basis node triggered next, and the parameter correlation type corresponding to the connection relationship is labeled, and the parameter correlation type is device parameter correlation or environment parameter correlation; The conduction directions of the basis nodes corresponding to all risk evolution paths are summarized, the conduction direction frequency corresponding to each basis node combination is counted, the contradictory conduction direction description with a conduction direction frequency lower than a preset proportion is removed, and the conduction direction consistent with most risk evolution paths is retained; According to the retained conduction direction, the trigger logic of all basis nodes is sorted in combination with the parameter correlation type, a basis node trigger sequence table is established, and a conduction sequence between the basis nodes is formed, the conduction sequence is consistent with the trigger sequence in the risk evolution path, and the basis node trigger sequence table includes node number, previous node number, next node number and parameter correlation type.

5. The substation switching operation risk prediction method based on big data analysis according to claim 1, characterized in that, The risk conduction strength analysis of the nodes and the conduction relationship in the switching operation risk association network is performed, and the risk conduction weight description corresponding to each association link is generated, including: The actual occurrence frequency data corresponding to each node in the switching operation risk association network in the historical operation risk information is extracted, the occurrence frequency data is statistically classified in layers, and a multi-period frequency distribution table is formed, and the occurrence frequency data is the number of times that the operation link unit or the risk factor corresponding to the node appears in the historical operation; The correlation degree of the occurrence frequency data corresponding to each node and the number of times of historical risk events is analyzed, the ratio of the number of times of risk events in a preset time window after the node appears to the total number of times of the node is calculated as the correlation degree, and the time variation trend of the correlation degree is analyzed in combination with the multi-period frequency distribution table; According to the correlation degree and the time variation trend of the correlation degree, a node importance coefficient is allocated to each node, the calculation basis and the time trend analysis result of the node importance coefficient are recorded, the higher the correlation degree and the more stable the time variation trend, the greater the node importance coefficient; The historical trigger success rate corresponding to each conduction connection relationship in the switching operation risk association network is analyzed, all trigger records corresponding to the conduction connection relationship in the historical operation risk information are queried, the trigger records include trigger condition satisfaction, actual trigger of the next node and parameter change data at the time of triggering, the number of valid records of the trigger condition satisfaction and the successful triggering of the next node is counted, and the total number of records of the trigger condition satisfaction is counted, and the ratio of the number of valid records to the total number of records is calculated as the trigger success rate. The triggering success rate of the conduction connection relationship is combined with the node importance coefficient to calculate a conduction strength value of each associated link, and the conduction strength value is a product of the node importance coefficient of the previous node and the triggering success rate of the corresponding conduction connection relationship; meanwhile, the conduction strength value is fine-tuned with reference to a ratio of a parameter change data amplitude to a historical average parameter change amplitude in the trigger record, and the greater the parameter change data amplitude ratio, the greater the fine-tuning amplitude; The conduction strength value is normalized, the weight level is divided according to the normalized result, and a risk conduction weight description corresponding to each associated link is generated, and the risk conduction weight description includes a weight value, a weight level identifier, a calculation basis and fine-tuning instructions.

6. The substation switching operation risk prediction method based on big data analysis according to claim 5, characterized in that, The triggering success rate of the conduction connection relationship is combined with the node importance coefficient to calculate a conduction strength value of each associated link, and the conduction strength value is a product of the node importance coefficient of the previous node and the triggering success rate of the corresponding conduction connection relationship; meanwhile, the conduction strength value is fine-tuned with reference to a ratio of a parameter change data amplitude to a historical average parameter change amplitude in the trigger record, and the greater the parameter change data amplitude ratio, the greater the fine-tuning amplitude; The node numbers of the previous node and the next node corresponding to each associated link are extracted, the node attribute tags in the reverse power operation risk association network are queried through the node numbers, and the node importance coefficient of the previous node and the calculation basis are obtained; All trigger records of the conduction connection relationship corresponding to the associated link in the historical operation risk information are queried, the trigger records are sorted in chronological order, and each trigger record includes a trigger condition satisfaction, an actual trigger condition of the next node, device parameter data at the time of triggering, environmental parameter data and a trigger result description; The number of valid records in which the trigger condition is satisfied and the next node is successfully triggered is counted, the normal range deviation values of the device parameter data and the environmental parameter data in the valid records are marked, the total number of records in which the trigger condition is satisfied is counted, and the number of records in which the trigger fails due to excessive parameter deviation is distinguished from the total number of records; The ratio of the number of valid records to the total number of records is calculated to obtain the triggering success rate of the conduction connection relationship, the average value of the normal range deviation values of the device parameter data and the environmental parameter data in the valid records is calculated as a parameter deviation reference value, and the average value of the parameter deviation values in all valid records is calculated as a historical average parameter deviation; The node importance coefficient of the previous node is multiplied by the triggering success rate to obtain an initial conduction strength value, and the initial conduction strength value is fine-tuned according to the ratio of the parameter deviation reference value to the historical average parameter deviation, the smaller the parameter deviation reference value ratio, the greater the fine-tuned conduction strength value, and the greater the parameter deviation reference value ratio, the smaller the fine-tuned conduction strength value; The fine-tuned conduction strength value is range-verified, compared with a preset reasonable conduction strength range, and abnormal values exceeding the reasonable range are eliminated, the causes of the abnormal values are analyzed, if the trigger record data is abnormal, the trigger record is re-counted, if the calculation logic is deviated, the calculation logic is corrected, and the conduction strength values conforming to the actual risk conduction logic are retained; The retained conduction strength value is taken as a final conduction strength value of each associated link, the calculation process of the final conduction strength value, the parameter deviation reference value ratio and the range verification result are recorded, and the final conduction strength value directly reflects the influence of the associated link in the risk conduction process.

7. The substation switching operation risk prediction method based on big data analysis according to claim 1, characterized in that, The risk transmission weight description is combined with real-time execution information of the current switching operation, a risk evolution path in the switching operation risk association network is deduced, and a risk prediction result of the substation switching operation is output, including: The real-time execution information of the current switching operation is analyzed, the operation link unit currently being executed, the corresponding device operation state information, the environment state information and the execution progress data are extracted, and real-time collection time stamps are added to the device operation state information and the environment state information; The identifier of the operation link unit currently being executed is matched with the operation link identifier of the operation link node in the switching operation risk association network, the corresponding target node is determined, the attribute label, the initial value of the association strength and the association basis record of the target node are obtained; Based on the risk transmission weight description, starting from the target node, according to the transmission path topology in the switching operation risk association network, all transmission paths starting from the target node are traversed, and the node sequence, the transmission connection relationship and the corresponding risk transmission weight description contained in each transmission path are recorded; According to the current device operation state information, the environment state information and the real-time collection time stamp, the state snapshot reference identifier in the attribute label of the subsequent node on each transmission path is compared with the corresponding historical state data, whether the transmission trigger condition of the subsequent node on each transmission path is met is judged, and the subsequent node and the corresponding judgment basis that meet the trigger condition are recorded; The risk level of the transmission path that meets the trigger condition is evaluated, the risk level evaluation combines the weight values in the risk transmission weight description of each associated link in the path for accumulated calculation, and simultaneously refers to the initial value of the association strength of the node in the transmission path, the greater the accumulated result is, the higher the risk level is; The transmission path whose risk level evaluation result exceeds the preset risk threshold is selected as the risk evolution path, and the key nodes and the trigger conditions in the risk evolution path are marked, the key nodes are the nodes corresponding to the associated link with the highest weight value; The risk factors, the possible risk event types, the key node information and the trigger conditions corresponding to all risk evolution paths are integrated to generate the risk prediction result of the substation switching operation, including the risk occurrence probability description, the risk influence range description, the risk trigger node information and the risk prevention and control suggestion, the risk occurrence probability description is determined based on the ratio of the number of risk evolution paths to the total number of transmission paths, and the risk influence range description is determined based on the device range and the operation range involved in the nodes in the risk evolution path.

8. The substation switching operation risk prediction method based on big data analysis according to claim 7, characterized in that, The trigger condition of the subsequent node on each transmission path is judged according to the current device operation state information and the environment state information, including: The transmission trigger condition description of the subsequent node on each transmission path is extracted, the transmission trigger condition description includes the state range of the device operation parameter, the state range of the environment parameter, the specification requirement of the operation execution and the parameter deviation allowable range; Extract actual parameter data corresponding to the conduction trigger condition description from the real-time execution information of the current switching operation, the actual parameter data including the actual value of the current device operating parameter, the actual value of the environmental parameter, the actual state of the operation execution, and the parameter collection time, which is consistent with the real-time collection time stamp; Compare the actual parameter data with the state range or specification requirement in the conduction trigger condition description, compare the actual value of the device operating parameter with the actual value of the environmental parameter corresponding to the state range, compare the actual state of the operation execution with the specification requirement of the operation execution, and record the deviation of each actual parameter data and the corresponding requirement; Statistically count the number of parameter items meeting the requirements in the conduction trigger condition of the subsequent node on each conduction path, the parameter items meeting the requirements being the items within the state range or specification requirement, and statistically count the total number of parameter items included in the conduction trigger condition, and distinguish between the key parameter items and the non-key parameter items in the total parameter items; Calculate the ratio of the number of parameter items meeting the requirements to the total number of parameter items to obtain the trigger condition satisfaction degree, and separately calculate the ratio of the number of key parameter items meeting the requirements to the total number of key parameter items as the key parameter satisfaction degree if there are key parameter items; Set the trigger condition satisfaction degree threshold and the key parameter satisfaction degree threshold, and determine that the trigger condition of the subsequent node on the conduction path is met when the trigger condition satisfaction degree reaches or exceeds the trigger condition satisfaction degree threshold and the key parameter satisfaction degree reaches or exceeds the key parameter satisfaction degree threshold; When the trigger condition satisfaction degree is lower than the trigger condition satisfaction degree threshold or the key parameter satisfaction degree is lower than the key parameter satisfaction degree threshold, it is determined that the trigger condition of the subsequent node on the conduction path is not met, and the parameter items that do not meet the requirements, the parameter item types, the deviation values and deviation reason analysis of the actual parameter data and the requirements are recorded, and the parameter item types are divided into key parameter items and non-key parameter items. 9.The substation switching operation risk prediction method based on big data analysis of claim 2, wherein, The risk factors associated with each operation link unit are screened from the device operating information, environmental influence information, and historical operation risk information in the multi-source associated information to form a risk factor list corresponding to each operation link unit, including: Classify and organize the device operating information, divide the device operating information into multiple device state data sets according to the device type, each device state data set containing the operating parameter change information, fault record information, maintenance record information, and parameter abnormal alarm record of the corresponding device, and adding a device unique identifier to each device state data set; Dimensionally split the environmental influence information into environmental data of three dimensions of natural environment information, field operation environment information, and external interference environment information, each dimension of environmental data containing data collection time, data value, and data change trend record; Structurally process the historical operation risk information, classify the historical risk events according to the risk event type, extract the associated device information, associated environment information, associated operation link information, risk occurrence time, risk influence range, and risk processing result corresponding to each historical risk event, and establish a historical risk event index library; For each operation link unit, the device state data corresponding to the device associated identifier in the device state data set is matched, the running parameter change information and parameter abnormal alarm record in the device state data are analyzed, the device class risk factors that may affect the execution of the operation link unit are determined, and the device class risk factors include parameter abnormal risk, device failure risk and maintenance missing risk; The environmental data overlapping with the execution time of the operation link unit is screened from the environmental data in three dimensions, the extreme value record in the natural environment information, the non-standard record in the field operation environment information and the interference intensity record in the external interference environment information are analyzed, the environmental class risk factors that may affect the execution of the operation link unit are determined, and the environmental class risk factors include natural environment abnormal risk, operation environment non-standard risk and external interference risk; The historical risk events associated with the operation link information and the action type consistent with the current operation link unit are screened from the historical risk event index library, the risk factors, associated device information and associated environment information corresponding to these historical risk events are extracted, the device associated identifier of the current operation link unit is compared with the associated device information of the historical risk events, and the historical risk factors matched with the devices are screened as the historical associated risk factors; The device class risk factors, the environmental class risk factors and the historical associated risk factors are combined, the consistency of the description of the risk factors is compared, the completely repeated risk factors are removed, the risk type identifier and the associated data source identifier are added to the retained risk factors, the risk type identifier is divided into device class risk identifier, environmental class risk identifier and historical associated class risk identifier, the risk factor list corresponding to each operation link unit is formed, the completeness of the risk factor list is verified, and the missing risk factor type is supplemented.

10. The substation switching operation risk prediction method based on big data analysis according to claim 3, characterized in that, The attribute label is added to each node in the switching operation risk association network, including: All basic nodes in the switching operation risk association network are traversed, the operation link identifier, risk factor identifier and trigger direction identifier contained in each basic node are read, the core information corresponding to each basic node is identified, and the core information is the operation link identifier or risk factor identifier; According to the type of the core information, the node type identifier of each basic node is allocated, when the core information is the operation link identifier, the node type identifier is set as the operation link node identifier, and the operation action description corresponding to the operation link identifier is recorded; when the core information is the risk factor identifier, the node type identifier is set as the risk factor node identifier, and the risk type description corresponding to the risk factor identifier is recorded; The association record corresponding to each basic node is extracted from the historical operation risk information, the association record is the number of times that the basic node and other nodes appear together in the historical risk event, the association record is weighted and counted according to the severity of the historical risk event, and the higher the severity of the historical risk event, the greater the weight of the corresponding association record. A ratio of a weighted association record quantity of each base node to an average weighted association record quantity of all nodes is calculated to obtain an association frequency coefficient, and a probability of the base node in all historical risk events is calculated, and the probability is a ratio of a number of historical risk events in which the base node appears to a total number of historical risk events; An initial value of an association strength of each base node is determined according to the association frequency coefficient and the probability, the higher the association frequency coefficient and the greater the probability, the greater the initial value of the association strength, and a corresponding relationship table of the association frequency coefficient, the probability and the initial value of the association strength is established; A node type identifier, the initial value of the association strength, core information description, the association frequency coefficient and the probability are integrated into an attribute label in a unified format, the attribute label and the corresponding base node are bound through a node number, and each node number uniquely corresponds to an attribute label through node number uniqueness verification; The bound base node is subjected to attribute verification, and whether the node type identifier and the core information in the attribute label are consistent, whether a calculation process of the initial value of the association strength is complete, and whether the association frequency coefficient and the probability are reasonable are checked, and an abnormal situation in which the attribute label and the node core information do not match, a calculation process is missing or a value is unreasonable is corrected, and a final node attribute label set is formed.

Images