Intelligent power transformation and distribution auxiliary data system based on multi-dimensional data model and abnormal event extension method

Through the intelligent transformer distribution auxiliary data system and abnormal event expansion method based on multi-dimensional data model, the problem of the degradation of the recognition ability of the intelligent transformer distribution system in the face of complex environments is solved, and the accuracy of abnormal detection and system reliability are improved. , reduces false positive rates, optimizes equipment maintenance and system evolution.

CN118944293AInactive Publication Date: 2024-11-12STATE GRID CORPORATION OF CHINA +2

Patent Information

Application Number
CN202411003118.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Intelligent substation distribution systems may lose their ability to identify new anomalies in complex environments, resulting in lagging detection and response, high false alarm rates and missed alarm rates, and are unable to effectively deal with unknown risks such as extreme weather, equipment aging and external attacks. .

Method used

Adopt an intelligent transformer distribution auxiliary data system based on multidimensional data model, combining multidimensional data model and abnormal events through intelligent inspection, safety defense, environmental monitoring, lighting control, water supply and drainage monitoring, fire fire protection, emergency lighting and online monitoring subsystem Extended methods, monitor and analyze data in real time, calculate subsystem correlation, label and record exception events to enrich exception models.

Benefits of technology

Improve the system's ability to identify various abnormal situations, reduce false alarms and missed alarm rates, enhance system reliability, provide historical data for equipment maintenance and optimization, extend equipment service life, and promote continuous improvement and evolution of the system. .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118944293A_ABST
    Figure CN118944293A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent power transformation and distribution auxiliary data system based on a multi-dimensional data model and an abnormal event extension method. The power transformation and distribution auxiliary data system comprises an intelligent inspection subsystem; a safety defense subsystem; an environment monitoring subsystem; a light control subsystem; a water supply and drainage monitoring subsystem; a fire fighting subsystem; an emergency lighting subsystem; and an online monitoring subsystem. The abnormal event extension method comprises the following steps: pre-establishing an abnormal model for storing typical abnormal events; extracting abnormal data of each subsystem; judging whether the abnormal data are matched with typical abnormal events stored in the abnormal model or not, if yes, outputting an alarm, and if not, marking the abnormal data and recording the abnormal data as marked abnormal events; and adding the marked abnormal event into the abnormal model as a newly added typical abnormal event. According to the invention, effective early warning and response can be carried out, rapid response can be ensured in case of sudden abnormal conditions, and stable operation of the system is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of transformer substation auxiliary control, and in particular to an intelligent transformer and distribution auxiliary data system based on a multi-dimensional data model. Background Art

[0002] In intelligent power distribution systems, anomaly detection and processing are important links to ensure system stability and security. In order to effectively deal with various abnormal situations in system operation, an accurate abnormal alarm system is necessary. With the increasing complexity of power distribution systems, the abnormal situations that may occur in system operation have become more diverse. Traditional abnormal models may only cover some common abnormal events, but in actual operation, many abnormalities that have not been recorded or processed will be encountered. These unrecorded abnormalities may be difficult to be accurately identified and processed by existing models due to their particularity or complexity. The requirement for the development of intelligent systems is to constantly adapt to new operating environments and challenges. If they cannot be updated in time, they may gradually lose the ability to identify new abnormalities, resulting in delayed detection and response. In addition, the accuracy of anomaly detection is directly related to the reliability of the system and the trust of users. If the existing model cannot fully cover all abnormal situations, there may be a high false alarm rate, that is, normal situations are misjudged as abnormalities, or a false alarm rate, that is, real abnormalities are ignored, which not only affects the operating efficiency of the system, but also may cause serious safety problems.

[0003] In summary, the power transformation and distribution system may face various unknown risks in actual operation, including extreme weather, equipment aging, external attacks, etc. If these unknown risks are not identified and handled in time, they may have a serious impact on the system. Therefore, it is necessary to record and analyze these unknown risks, conduct effective early warning and response, and ensure that they can respond quickly when faced with sudden abnormal situations to ensure the stable operation of the system. Summary of the invention

[0004] In order to solve the problem that the power transformation and distribution system may face various unknown risks in actual operation, in terms of coping with complex environments, improving abnormality detection and response capabilities, and ensuring stable operation of the system, the present invention provides an intelligent power transformation and distribution auxiliary data system based on a multidimensional data model and an abnormal event expansion method.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] On the one hand, the present invention discloses an intelligent power substation auxiliary data system based on a multidimensional data model, including: an intelligent patrol subsystem: used to access monitoring information of various patrol equipment to realize centralized collection and analysis of substation patrol data;

[0007] Security defense subsystem: used to monitor the operation of access controllers, electronic fences, infrared dual-detection detectors, infrared beam detectors, and sound and light alarms;

[0008] Environmental monitoring subsystem: used to access monitoring information of temperature and humidity sensors, wind speed sensors, SF6 sensors, and protocol converters;

[0009] Lighting control subsystem: used to realize lighting control;

[0010] Water supply and drainage monitoring subsystem: used to monitor the pressure, flow rate and flow of the water supply and drainage system, and to monitor and control the water conditions in the station by connecting to water immersion sensors and water level sensors to sense information;

[0011] Fire fighting subsystem: used to detect fire in different areas;

[0012] Emergency lighting subsystem: used to ensure emergency lighting needs when AC power fails, and linked with the fire fighting subsystem;

[0013] Online monitoring subsystem: used for online monitoring of transformers, GIS / HGIS, circuit breakers, air-insulated switchgear, gas-insulated switchgear, and lightning arrester data, to achieve real-time perception and real-time monitoring of equipment monitoring status.

[0014] Furthermore: the system front-end equipment of the intelligent patrol subsystem includes panoramic cameras, high-definition gun cameras, high-definition ball cameras, high-definition explosion-proof hemispherical cameras, pan-tilt cameras, and infrared thermal imaging cameras.

[0015] Furthermore: the system front-end equipment of the lighting control subsystem includes: a lighting control module, a light illumination sensor, an infrared sensing sensor and a lighting circuit. By collecting the on-off status of the lighting circuit, the light illumination sensor and the infrared sensing sensor information, remote control, area control, light illumination control and infrared sensing control are realized.

[0016] Furthermore: the fire fighting subsystem includes a fire fighting information transmission control unit, an automatic fire alarm system, and an analog transmitter to realize the collection, transmission and linkage control of fire alarm information within the station.

[0017] Furthermore: the emergency lighting subsystem includes an emergency lighting controller, emergency lighting fixtures, and emergency sign lamps.

[0018] Furthermore: the online monitoring subsystem includes collecting oil level monitoring, dissolved gas monitoring in transformer oil, core clamp grounding current monitoring, arrester leakage current monitoring, insulating gas density monitoring, and switch contact temperature online monitoring data to achieve real-time perception and real-time monitoring of equipment monitoring status.

[0019] On the other hand, the present invention discloses an abnormal event expansion method applicable to power transformation and distribution auxiliary use, comprising the following steps:

[0020] Pre-establish anomaly models for storing typical abnormal events;

[0021] By calculating the correlation of each subsystem, the abnormal data of each subsystem is extracted;

[0022] Determine whether the abnormal data matches the typical abnormal events stored in the abnormal model. If they match, an alarm is output. If they do not match, the abnormal data is marked and recorded as a marked abnormal event. The marking process includes:

[0023] A. Abnormal data appears in a subsystem caused by a single event, and the abnormal event is recorded as a marked abnormal event;

[0024] B. The number of abnormalities in multiple subsystems caused by a single event is recorded as a marked abnormal event;

[0025] C. Abnormal data appears in a subsystem caused by multiple abnormal events, and the multiple abnormal events are recorded as a marked abnormal event;

[0026] D. Multiple subsystems triggered by multiple abnormal events have abnormal data, and the multiple abnormal events are recorded as a marked abnormal event;

[0027] The marked abnormal events are added to the abnormal model as new typical abnormal events.

[0028] The marked abnormal events are added to the abnormal model as typical abnormal events.

[0029] Further: the calculation of the correlation degree of each subsystem includes:

[0030] Divide the data of each subsystem into fixed time windows;

[0031] Aggregate the data in each time window to generate time series data;

[0032] Assume that there are n subsystems, and represent the data generated by each subsystem at T time points as an n×T time series data matrix D; each row in the matrix D represents the data of a subsystem at different time points, and each column represents the data of all subsystems at the same time point;

[0033] Using the time series data matrix D, calculate the mutual information I(X i ;X j) to measure their dependence: Using the time series data matrix D, calculate the correlation coefficient between each pair of subsystem time series measure their linear correlation;

[0034] The calculated mutual information and correlation coefficient are stored in the association matrix respectively, and the association matrix I and ρ are obtained.

[0035] Further: the extraction of abnormal data of each subsystem includes:

[0036] For mutual information, set the upper threshold T I and the lower threshold L I , for the correlation coefficient, set the upper threshold T ρ and the lower threshold L ρ :

[0037] For each pair of subsystems X i and X j , if I(X i ;X j )>T I or I(X i ;X j )<L I , it is marked as abnormal;

[0038] For each pair of subsystems X i and X j ,if or It is marked as abnormal;

[0039] X according to the marked exception i and X j , extract the subsystem data points within the corresponding time window as abnormal data.

[0040] Compared with the prior art, the present invention has the following technical advances:

[0041] The more abnormal events the present invention marks, the richer the typical abnormal events in the abnormal model. More marked events mean that the system has a stronger ability to identify various abnormal situations, can more accurately judge and classify new abnormal events, and can more accurately distinguish between normal and abnormal situations, reduce false alarm rates and missed alarm rates, and improve the reliability of the system. At the same time, the rich records of abnormal events provide historical data for subsequent maintenance, help identify long-term trends and potential problems, and through the analysis of abnormal events, optimize equipment maintenance plans, reduce unnecessary maintenance, and extend the service life of equipment. More marked events also provide a basis for continuous learning and optimization of the model, so that the system can continuously improve its own performance and adaptability. By marking and recording new abnormal events, a feedback loop is formed to promote the continuous improvement and evolution of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0043] In the attached picture:

[0044] Figure 1 is a system structure diagram of Embodiment 1 of the present invention;

[0045] Figure 2 is a flow chart of Embodiment 2 of the present invention; DETAILED DESCRIPTION

[0046] The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0047] Example 1

[0048] like Figure 1 As shown, the present invention discloses an intelligent power transformation and distribution auxiliary data system based on a multidimensional data model, comprising:

[0049] The intelligent patrol subsystem, including the system front-end equipment mainly includes panoramic cameras, high-definition gun cameras, high-definition ball cameras, high-definition explosion-proof hemispheric cameras, pan-tilt cameras, infrared thermal imaging cameras, etc. The high-definition cameras must have night vision functions.

[0050] a) The configuration of the front-end camera of the intelligent patrol subsystem follows the following principles:

[0051] 1) The main transformer (including the regulating transformer) and high-resistance transformer are equipped with infrared temperature measurement cameras to realize automatic temperature measurement and inspection of the main body, casing, lead head, drainage line, oil pillow and end screen;

[0052] 2) For meters that cannot achieve remote data transmission (such as casing oil level meters), configure cameras to collect image data;

[0053] 3) For meters that can achieve remote data transmission (oil temperature meter, winding temperature meter, oil pressure meter, etc.), cameras are not configured for image data acquisition;

[0054] 4) There are no cameras in the control cabinet and terminal box to monitor the internal environment;

[0055] 5) When configuring a dual-spectrum infrared temperature measurement camera, the number of visible light cameras should be reduced;

[0056] 6) When configuring a patrol robot, the patrol area that the robot can cover should be reduced by the number of infrared temperature measurement cameras and visible light cameras;

[0057] 7) Visible light cameras should preferably be digital high-definition cameras with a resolution of no less than 1080P.

[0058] b) A fixed camera is installed at the entrance of the substation gate to monitor the license plates of vehicles entering and exiting the gate. At the same time, a fixed camera is deployed for routine monitoring of the gate entrance and exit, as well as environmental status analysis and personnel behavior analysis.

[0059] c) Each corner of the substation wall is equipped with a dome camera for perimeter anti-intrusion monitoring. One dome camera is added for every 70-meter length of a single wall.

[0060] d) Outdoor substations with voltage levels of 220kV and above are equipped with 1 to 3 panoramic cameras, and outdoor substations with voltage levels of 110kV and below are equipped with 1 to 2 panoramic cameras for station environment monitoring and panoramic monitoring of equipment areas.

[0061] e) The configuration principles of the main transformer / high-voltage shunt reactor camera are as follows:

[0062] 1) 500kV main transformer / high-voltage reactor: 1 unit on the top, each main transformer / high-voltage reactor is equipped with 2 to 5 dome cameras or PTZ cameras. Each main transformer / high-voltage reactor is equipped with 1 thermal imaging temperature measurement camera.

[0063] 2) 220kV indoor: 1 unit on the top, 3 to 5 dome cameras or PTZ cameras for each main transformer, and 1 thermal imaging temperature measuring camera for each main transformer.

[0064] 3) 220kV outdoor main transformer: Each main transformer is equipped with 1 to 2 dome cameras or pan-tilt cameras, and each main transformer is equipped with 1 thermal imaging temperature measurement camera.

[0065] 4) 110kV (66kV) indoor main transformer: 1 on each side and top, each main transformer is equipped with 3 to 5 dome cameras or pan-tilt cameras, and each main transformer is equipped with 1 thermal imaging temperature measurement camera.

[0066] 5) 110kV (66kV) outdoor main transformer: Each main transformer is equipped with 1 to 2 dome cameras or pan-tilt cameras, and each main transformer is preferably equipped with 1 thermal imaging temperature measurement camera.

[0067] 6) 35kV indoor / outdoor main transformer: Each main transformer is equipped with 1 dome camera or PTZ camera.

[0068] f) The camera configuration principles for the power distribution device area are as follows:

[0069] 1) 500kV power distribution equipment area: 1-2 dome cameras or PTZ cameras are installed every 3 bays;

[0070] 2) 220kV, 110kV power distribution equipment area: 1-2 dome cameras or PTZ cameras are configured for every 2 bays;

[0071] 3) 35kV outdoor power distribution equipment area: 1 to 2 dome cameras or pan-tilt cameras are configured every 2 intervals.

[0072] g) The camera configuration principles for the main control room, communication room and other rooms of 220kV~500kV substations are as follows:

[0073] 1) Main monitoring room: 1 to 2 dome cameras in each room;

[0074] 2) Relay room, secondary equipment room and communication room: 1 to 2 dome cameras in each room;

[0075] 3) Switchgear room: Each room is equipped with 2 ball cameras for monitoring the indoor panoramic view, arranged diagonally; every 8 switchgears (4 on the left and 4 on the right) are equipped with 1 ball camera for remote viewing function;

[0076] 4) Capacitor / reactor room: Each room is equipped with 2 to 3 dome cameras; 1 to 2 thermal imaging temperature measurement cameras can be optionally installed as needed;

[0077] 5) Station transformer / ground transformer room: each room is equipped with 1 to 2 dome cameras;

[0078] 6) Fire pump room: Each room is equipped with a dome camera;

[0079] 7) Cable mezzanine: A spherical camera should be installed at each water collection well, cable entry and exit holes, and entrance and exit.

[0080] The security defense subsystem uses security monitoring terminals, electronic fences, infrared beam detectors, infrared dual-detection detectors, access control controllers, sound and light alarms, emergency alarm buttons, burglar alarm controllers, etc. to achieve intrusion detection, alarms, and access control.

[0081] The front-end hardware configuration of the security defense subsystem is as follows:

[0082] a) The surrounding walls are equipped with high-voltage pulse electronic fences. A dual-zone host or a single-zone host is selected according to the actual situation on site. A defense zone of 100 meters is adopted, and a 6-wire system is adopted. According to local public security prevention requirements, a defense zone of less than 70 meters can be adopted. The electronic fence meets the relevant design and technical requirements of GB7946 and has obtained certification from the local public security department.

[0083] b) Use 1 to 2 pairs of infrared beams above the door.

[0084] c) Infrared dual-detection detectors shall be installed as needed at the entrances and exits of buildings and at the windows on the first floor.

[0085] d) Infrared dual-detection detectors should be installed at the entrances and exits of the power distribution device room, secondary equipment room, etc.; long-distance infrared dual-detection detectors should be installed for some GIS rooms with long corridors.

[0086] e) Access control is installed at the substation entrance gate, building entrances and exits, secondary equipment room, high-voltage room, communication room, battery room, GIS room, and capacitor room.

[0087] f) A video intercom doorbell shall be installed at the main entrance of 220kV and above substations.

[0088] g) For substations that need to remotely transmit alarm information to the 110 networked alarm center, one anti-theft alarm controller can be configured.

[0089] h) A 500kV substation is equipped with one emergency alarm button, and a substation with a voltage level of 220kV and below should be equipped with one emergency alarm button.

[0090] i) Sound and light alarms should be installed around the walls, and should be installed in the secondary equipment room, switch room, and GIS room.

[0091] j) All types of front-end hardware mentioned above shall not be installed on the aluminum-magnesium-manganese exterior wall panels of the steel structure substation using screws.

[0092] The environmental monitoring subsystem uses environmental monitoring terminals, air conditioning protocol converters, dehumidifier protocol converters, fan protocol converters, temperature and humidity sensors, micro-meteorological sensors, SF6 sensors and other equipment to realize real-time collection, processing and uploading of substation operating environment data.

[0093] The front-end hardware configuration of the environmental monitoring subsystem is as follows:

[0094] a) In important equipment rooms such as secondary equipment rooms, capacitor rooms, station transformer rooms, power distribution device rooms, battery rooms, independent communication rooms and prefabricated cabins, one temperature and humidity sensor should be installed for every 20 square meters in a single room.

[0095] b) The outdoor station is equipped with one wind speed sensor, while the indoor station is not equipped with one.

[0096] c) GIS room and other power distribution room containing SF6 equipment should be equipped with one SF6 leakage sensor per compartment. SF6 circuit breaker switch cabinet room should be equipped with one SF6 leakage sensor per 30 square meters.

[0097] d) The air conditioning system has control and communication functions. Each split air conditioner can be equipped with one air conditioning protocol converter; each multi-split air conditioning system can be equipped with one air conditioning protocol converter.

[0098] e) The fan system has communication function; the fans (other ordinary fans not used for fire fighting except axial flow fans) installed in equipment rooms with large heat generation such as transformer room and reactive power compensation device room have variable frequency speed control function. Each fan can be equipped with one fan protocol converter.

[0099] f) The dehumidifier system has control and communication functions. Each dehumidifier can be equipped with a dehumidifier protocol converter.

[0100] g) Outdoor venues should be equipped with an integrated micro-meteorological subsystem to collect data such as outdoor temperature, humidity, wind speed, wind direction, air pressure, and rainfall.

[0101] The lighting control subsystem includes a lighting control module, a light illumination sensor, an infrared sensor, and a lighting circuit. By collecting information such as the on / off status of the lighting circuit and sensors, it can realize various lighting circuit control methods such as remote control, area control, light illumination control, and infrared control.

[0102] Indoor corridor / passage lighting should be equipped with infrared sensors to realize the function of "lights on when people come and lights off when people leave" by detecting the infrared rays of the human body. The number of sensors is determined according to the number of lamps.

[0103] Auxiliary lights are configured for sites within the station that require additional lighting (such as outdoor fixed cameras and indoor cable layer cameras). Auxiliary lights can be sensor-based fill lights, which can be controlled based on the illumination conditions.

[0104] Lighting control modules are installed in the outdoor power distribution device area and important secondary equipment rooms such as the secondary equipment room and battery room. The number of lighting control modules is determined according to the actual number of lighting fixtures controlled.

[0105] The water supply and drainage monitoring subsystem realizes

[0106] Collect information such as pump operation status, fault alarm, and water conditions in the station, and realize the following functions:

[0107] a) Support remote start / stop control and maintenance tagging of water pumps.

[0108] b) Support water level threshold alarm configuration and alarm mode setting.

[0109] c) Support automatic control of water pump start when water level alarm is triggered in water collection well, and automatic stop when alarm is restored. The front-end equipment of water supply and drainage monitoring can communicate with water supply and drainage monitoring terminal by RS485, I / O and other interfaces.

[0110] The device configuration follows the following principles:

[0111] a) Small wall-mounted water supply and drainage monitoring terminals should be configured according to the equipment area; centralized water supply and drainage monitoring terminals should be configured according to the number and location of substation environmental monitoring front-end equipment. Each substation should be equipped with at least one set of water supply and drainage monitoring terminals (including I / O and RS485 interface type devices).

[0112] b) Water immersion sensors should be installed in cable layers, cable trenches and other cable concentration areas. Water immersion sensors should be placed in places where water is likely to accumulate in the cable trench, with one sensor installed every 60 meters or each section in the cable trench.

[0113] c) Each water collection well is equipped with a water level sensor to achieve accurate measurement of the water level within the measuring range.

[0114] d) Each group of water pumps is equipped with a water pump controller according to signal acquisition or control requirements; each water pump can be controlled independently.

[0115] e) Each water supply pipeline should be equipped with a digital remote meter.

[0116] The fire fighting subsystem, including the fire information transmission control unit, automatic fire alarm system, analog transmitter and other equipment, cooperates with the automatic fire alarm system to realize the collection, transmission and linkage control of fire alarm information within the station.

[0117] a) It can receive early warning information of fire, fire alarm information, analog quantity acquisition information, status information of on-site fire fighting facilities and start fire extinguishing equipment.

[0118] b) It has fire emergency handling function.

[0119] c) The fire information transmission control unit has the functions of self-checking and alarming, receiving and sending fire alarm, analog quantity acquisition, status of controlled fire-fighting facilities, start-stop action feedback information, receiving remote control and starting and stopping the controlled fire-fighting facilities, and performing local settings and debugging through the human-machine interface.

[0120] The emergency lighting subsystem consists of emergency lighting controllers, emergency lighting fixtures, emergency sign lamps, etc. Its main function is to provide necessary illumination conditions for the safe evacuation of personnel and fire fighting and rescue operations in emergency situations such as fire.

[0121] The design follows the basic design principles of simple system architecture and simple control, including lighting layout, system power distribution, system control design under non-fire conditions, and system control design under fire conditions.

[0122] In a fire state, the relevant linkage between the emergency lighting subsystem and the fire fighting subsystem is completed by the fire alarm controller; in a non-fire state, the relevant linkage control is completed by the auxiliary control background.

[0123] Online monitoring subsystem, to monitor oil temperature and oil level, transformer oil dissolved gas, core clamp grounding current,

[0124] Functions include lightning arrester leakage current monitoring, insulating gas density monitoring, switch contact temperature measurement, etc.

[0125] a) Main transformer online monitoring device

[0126] 1) The transformer monitoring terminal is configured according to the number of main transformers in the substation. Each main transformer should be equipped with a transformer monitoring terminal.

[0127] 2) Oil-immersed transformers and reactors of 220 kV and above should be equipped with an online monitoring unit for dissolved gas in oil, which should be configured for each transformer (reactor).

[0128] 3) Transformers of 110kV and above should be equipped with an online monitoring device for core / clamp grounding current, which should be configured for each transformer.

[0129] 4) Transformers of 35kV and above are equipped with digital remote transmission meters for oil temperature and oil level.

[0130] 5) 110kV~220kV transformers are equipped with complete sets of neutral point arrester and digital remote leakage current meter.

[0131] b)GIS / HGIS online monitoring device

[0132] 1) GIS / HGIS monitoring terminals should be configured according to voltage levels.

[0133] 2) GIS / HGIS of 110kV and above shall be equipped with an insulating gas density remote transmission meter.

[0134] 3) GIS / HGIS with built-in lightning arresters of 110kV and above are equipped with digital remote leakage current meters.

[0135] c) Circuit breaker online monitoring device

[0136] 1) The meter data collection terminal should be configured according to the voltage level.

[0137] 2) Gas insulated circuit breakers of 35kV and above are equipped with an insulating gas density remote transmission meter.

[0138] d) Air-insulated switchgear (only used for high-current switchgear such as incoming line cabinets and sectionalizing cabinets)

[0139] 1) The contact temperature monitoring terminal should be configured according to the switch cabinet room.

[0140] 2) 10kV incoming line cabinet, section cabinet and other air-insulated switch cabinets are equipped with contact temperature monitoring.

[0141] e) Gas insulated switchgear

[0142] 1) The meter collection terminal should be configured according to the switch cabinet room.

[0143] 2) 35kV and 10kV gas insulated switchgear are equipped with an insulating gas density remote transmission meter.

[0144] f) Independent lightning arrester

[0145] 1) The meter data collection terminal should be configured according to the voltage level or connected to the equipment interval data collection terminal.

[0146] 2) For arresters of 110kV and above, a digital remote leakage current meter is installed.

[0147] Example 2

[0148] According to the idea of ​​the inventor's technical team, the sub-data structures / data sub-systems corresponding to each sub-system are integrated into one, for example, a data map / data space is constructed; then, only abnormal data or abnormal data fluctuations are focused on, and associated parameters are constructed for the abnormal data; then modeling is started:

[0149] A. Pre-modeling (pre-construct a subspace in the initially constructed data map / data space, and store the correlation of abnormal data caused by typical and conventional fault events in it). According to the pre-stored model, the background can directly prompt possible abnormal power events based on the current data correlation.

[0150] B. Label modeling. The actual power operation is complicated. The emergence of various power abnormal data cannot completely follow the pre-modeling corresponding to typical events. Label modeling is required at will. According to the current atypical abnormal data pattern and the actual power abnormal events obtained after manual analysis and processing (after the analysis and processing of the current power abnormality is completed and archived), the data is labeled, thereby realizing the expansion of a limited number of pre-stored typical models, which become more and more abundant with use.

[0151] Based on this, this embodiment discloses an intelligent power distribution auxiliary data system based on a multidimensional data model, such as Figure 2 As shown, the following steps are included:

[0152] S01: pre-establish an abnormal model for storing typical abnormal events;

[0153] S02: extracting abnormal data of each subsystem by calculating the correlation degree of each subsystem;

[0154] Specifically, calculating the correlation of each subsystem includes:

[0155] The data of each subsystem is divided according to fixed time windows (such as every minute, every hour), and the data is aggregated in each time window (such as calculating the average, maximum, minimum, etc.) to generate time series data.

[0156] Assuming there are n subsystems and the data length of each subsystem is T, these data can be represented as an n×T time series data matrix D. Each row in the matrix D represents the data of a subsystem at different time points, and each column represents the data of all subsystems at the same time point.

[0157] For two subsystems X i and X j , we first need to estimate the probability distribution p(x i , x j ) and their respective marginal probability distributions p(x i ) and p(x j ), specifically, using the time series data matrix D, calculate the mutual information between each pair of subsystem time series to measure their dependency. For the specific calculation process, the following algorithm can be used directly: Using the time series data matrix D, the correlation coefficient between each pair of subsystem time series is calculated to measure their linear correlation. For the specific calculation process, the following algorithm can be used directly: Among them, X i and X j is the time series data of the ith and ith subsystems, p(x i , x j ) is the joint probability density function, p(x i ) and p(x j ) is the marginal probability density function, Cov(X i , X j ) is the covariance, and They are X i and X j The standard deviation of .

[0158] By constructing time series data, standardized input data is provided, so that the correlation calculation between subsystems can be performed on a unified time scale.

[0159] The calculated mutual information and correlation coefficient are stored in the association matrix respectively, and the association matrix I and ρ are obtained.

[0160] Specifically, the abnormal data extracted from each subsystem includes:

[0161] For all elements in the mutual information matrix I, calculate the mean μ and standard deviation σ I, for all elements in the correlation coefficient matrix ρ, calculate the mean μ ρ and standard deviation σ ρ For mutual information, set the upper threshold T I and the lower threshold L I , for the correlation coefficient, set the upper threshold T ρ and the lower threshold L ρ , where the upper threshold T I , lower threshold L I , upper threshold T ρ and the lower threshold L ρ The mean plus or minus three standard deviations was used.

[0162] For each pair of subsystems X i and X j , if I(X i ;X j )>T I or I(X i ;X j )<L I , it is marked as abnormal;

[0163] For each pair of subsystems X i and X j ,if or It is marked as abnormal.

[0164] For the subsystem pairs X1 and X2 marked as abnormal, their original data points in the corresponding time window are extracted as abnormal data.

[0165] S03: Determine whether the abnormal data matches the typical abnormal events stored in the abnormal model. If they match, an alarm is output; if they do not match, the abnormal data is marked.

[0166] Specifically, a data algorithm can be constructed for automatic matching to determine whether the abnormal data matches the typical abnormal events stored in the abnormal model, including:

[0167] Use typical abnormal events stored in the abnormal model as training sets;

[0168] Assume that the training set contains k labeled typical abnormal event feature vectors F = {f1, f2, ..., f k}, the corresponding label is y = {y1, y2, ..., y k}, where y i ∈{0, 1} represents normal and abnormal.

[0169] Using the kernel function K(f i , f j ) maps the feature vector to a high-dimensional space:

[0170] Input the extracted abnormal data feature vector into the trained training model to obtain the classification result

[0171] If the classification result Then the abnormal data matches the typical abnormal event and an alarm is output;

[0172] If the classification result If the abnormal data does not match the typical abnormal event, the abnormal data will be marked.

[0173] It should be noted that the matching process can be performed in a variety of ways, and this embodiment only discloses a typical data algorithm. Abnormal events that cannot be automatically classified are submitted to experts for manual labeling and recorded as labeled abnormal events. The labeling process includes:

[0174] A. Abnormal data appears in a subsystem caused by a single event. The abnormal event is recorded as a marked abnormal event. For example:

[0175] Event: Windy weather.

[0176] Subsystem: Intelligent patrol subsystem.

[0177] Abnormal data: The surveillance image of the fixed camera at the entrance of the substation gate is abnormal.

[0178] Mark: The event in which strong wind causes abnormal monitoring images of the intelligent patrol subsystem camera is recorded as a marked abnormal event.

[0179] B. Abnormal data appears in multiple subsystems caused by a single event. The abnormal event is recorded as a marked abnormal event. Example:

[0180] Event: Heavy rain.

[0181] Subsystems: Environmental monitoring subsystem and water supply and drainage monitoring subsystem.

[0182] Abnormal data:

[0183] Environmental Monitoring Subsystem: Wind speed sensor reading is abnormal.

[0184] Water supply and drainage monitoring subsystem: water level sensor reading is abnormal.

[0185] Mark: The event in which heavy rain causes abnormal readings of the wind speed sensor of the environmental monitoring subsystem and the water level sensor of the water supply and drainage monitoring subsystem is recorded as a marked abnormal event.

[0186] C. Abnormal data appears in a subsystem caused by multiple abnormal events. The multiple abnormal events are recorded as a marked abnormal event. For example:

[0187] Event: High winds and lightning strikes.

[0188] Subsystem: Intelligent patrol subsystem.

[0189] Abnormal data: The surveillance image of the fixed camera at the entrance of the substation gate is abnormal.

[0190] Marking: Combine the events caused by strong wind and lightning strikes that cause abnormal monitoring images of the intelligent patrol subsystem camera and record them as a marked abnormal event.

[0191] D. Multiple subsystems triggered by multiple abnormal events have abnormal data, and the multiple abnormal events are recorded as a marked abnormal event. Example:

[0192] Event: Lightning strike and heavy rain.

[0193] Subsystems: Environmental monitoring subsystem and water supply and drainage monitoring subsystem.

[0194] Abnormal data:

[0195] Environmental monitoring subsystem: SF6 sensor reading is abnormal.

[0196] Water supply and drainage monitoring subsystem: The water sensor reading is abnormal.

[0197] Marking: The events caused by lightning strikes and heavy rain that cause abnormal readings of the SF6 sensor of the environmental monitoring subsystem and the water immersion sensor of the water supply and drainage monitoring subsystem are combined and recorded as a marked abnormal event.

[0198] These markers can accurately record the causes and impacts of abnormal events, improving the system's anomaly detection and response capabilities.

[0199] S04: Add the marked abnormal events to the abnormal model as new typical abnormal events.

[0200] As the complexity of the power transformation and distribution system increases, the associations and interactions between subsystems become more diverse. Expanding the anomaly model and storing more typical abnormal events by marking can make the system more accurate in identifying and detecting anomalies. The rich abnormal event library also provides a wide range of references, which can more accurately judge new abnormal situations and improve the reliability of anomaly detection.

[0201] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the claims of the present invention.

Claims

1. An intelligent power transformation and distribution auxiliary data system based on a multidimensional data model, characterized in that: include: Intelligent patrol subsystem: used to access monitoring information of various patrol equipment to realize centralized collection and analysis of substation patrol data; Security defense subsystem: used to monitor the operation of access controllers, electronic fences, infrared dual-detection detectors, infrared beam detectors, and sound and light alarms; Environmental monitoring subsystem: used to access monitoring information of temperature and humidity sensors, wind speed sensors, SF6 sensors, and protocol converters; Lighting control subsystem: used to realize lighting control; Water supply and drainage monitoring subsystem: used to monitor the pressure, flow rate and flow of the water supply and drainage system, and to monitor and control the water conditions in the station by connecting to water immersion sensors and water level sensors to sense information; Fire fighting subsystem: used to detect fire in different areas; Emergency lighting subsystem: used to ensure emergency lighting needs when AC power fails, and linked with the fire fighting subsystem; Online monitoring subsystem: used for online monitoring of transformers, GIS / HGIS, circuit breakers, air-insulated switchgear, gas-insulated switchgear, and lightning arrester data, to achieve real-time perception and real-time monitoring of equipment monitoring status.

2. The intelligent power transformation and distribution auxiliary data system based on a multidimensional data model according to claim 1 is characterized in that: The system front-end equipment of the intelligent patrol subsystem includes panoramic cameras, high-definition gun cameras, high-definition ball cameras, high-definition explosion-proof hemispherical cameras, pan-tilt cameras, and infrared thermal imaging cameras.

3. The intelligent power transformation and distribution auxiliary data system based on multidimensional data model according to claim 1 is characterized in that: The system front-end equipment of the lighting control subsystem includes: a lighting control module, a light illumination sensor, an infrared sensing sensor and a lighting circuit. By collecting the on-off status of the lighting circuit, the light illumination sensor and the infrared sensing sensor information, remote control, area control, light illumination control and infrared sensing control are realized.

4. The intelligent power transformation and distribution auxiliary data system based on multidimensional data model according to claim 1 is characterized in that: The fire fighting subsystem includes a fire fighting information transmission control unit, an automatic fire alarm system, and an analog transmitter to realize the collection, transmission, and linkage control of fire alarm information within the station.

5. The intelligent power transformation and distribution auxiliary data system based on multidimensional data model according to claim 1 is characterized in that: The emergency lighting subsystem includes an emergency lighting controller, emergency lighting fixtures, and emergency sign lamps.

6. The intelligent power transformation and distribution auxiliary data system based on multidimensional data model according to claim 1 is characterized in that: The online monitoring subsystem includes collecting oil level monitoring, transformer oil dissolved gas monitoring, core clamp grounding current monitoring, arrester leakage current monitoring, insulating gas density monitoring, and switch contact temperature online monitoring data to achieve real-time perception and real-time monitoring of equipment monitoring status.

7. A method for expanding abnormal events for auxiliary power transformation and distribution, characterized in that: The steps include: Pre-establish anomaly models for storing typical abnormal events; By calculating the correlation of each subsystem, the abnormal data of each subsystem is extracted; Determine whether the abnormal data matches the typical abnormal events stored in the abnormal model. If they match, an alarm is output. If they do not match, the abnormal data is marked and recorded as a marked abnormal event. The marking process includes: A. Abnormal data appears in a subsystem caused by a single event, and the abnormal event is recorded as a marked abnormal event; B. The number of abnormalities in multiple subsystems caused by a single event is recorded as a marked abnormal event; C. Abnormal data appears in a subsystem caused by multiple abnormal events, and the multiple abnormal events are recorded as a marked abnormal event; D. Multiple subsystems triggered by multiple abnormal events have abnormal data, and the multiple abnormal events are recorded as a marked abnormal event; The marked abnormal events are added to the abnormal model as new typical abnormal events.

8. The abnormal event expansion method applicable to power transformation and distribution auxiliary according to claim 7 is characterized in that: The calculation of the correlation degree of each subsystem includes: Divide the data of each subsystem into fixed time windows; Aggregate the data in each time window to generate time series data; Assume that there are n subsystems, and represent the data generated by each subsystem at T time points as an n×T time series data matrix D; each row in the matrix D represents the data of a subsystem at different time points, and each column represents the data of all subsystems at the same time point; Using the time series data matrix D, calculate the mutual information I(X i ;X j ) to measure their dependence: Using the time series data matrix D, calculate the correlation coefficient between each pair of subsystem time series measure their linear correlation; The calculated mutual information and correlation coefficient are stored in the association matrix respectively, and the association matrix I and ρ are obtained.

9. The abnormal event expansion method applicable to power transformation and distribution auxiliary according to claim 8 is characterized in that: The extracting of abnormal data of each subsystem includes: For mutual information, set the upper threshold T I and the lower threshold L I , for the correlation coefficient, set the upper threshold T ρ and the lower threshold L ρ : For each pair of subsystems X i and X j , if I(X i ;X j )>T I or I(X i ;X j ) <L I , it is marked as abnormal; For each pair of subsystems X i and X j ,if or It is marked as abnormal; X according to the marked exception i and X j , extract the subsystem data points within the corresponding time window as abnormal data.

Citation Information

Patent Citations

  • Intelligent identification and linkage method of alarm triggering event based on one-button alarm column

    CN109920218A

  • Abnormal intelligent inspection system and method for substation equipment

    CN110888373A

  • Intelligent power distribution room sensor multi-dimensional data fusion abnormal event detection method and system and computer readable storage medium

    CN114328075A

  • Auxiliary control method for transformer substation

    CN117526574A

  • Intelligent substation auxiliary control system

    CN118040877A

Cited By

  • Control method and device of intelligent lamp pole equipment compartment, electronic equipment and storage medium

    CN119576056A

  • Prefabricated cabin type charging integrated equipment and control method thereof

    CN121261397A