CN110349048A - Substation's multidimensional data runs interactive controlling platform and failure method of disposal - Google Patents
Substation's multidimensional data runs interactive controlling platform and failure method of disposal Download PDFInfo
- Publication number
- CN110349048A CN110349048A CN201910452000.8A CN201910452000A CN110349048A CN 110349048 A CN110349048 A CN 110349048A CN 201910452000 A CN201910452000 A CN 201910452000A CN 110349048 A CN110349048 A CN 110349048A
- Authority
- CN
- China
- Prior art keywords
- current break
- current
- data
- multidimensional data
- equipment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention discloses a kind of substation's multidimensional data operation interactive controlling platform and failure methods of disposal, when D5000 power scheduling monitor supervision platform monitors that current break occurs, the information of monitoring is analyzed, identify current break reason, wherein current break occurrence cause includes that accident occurs for system, normal service needs and dispatching running way adjustment, the current break reason short message of identification is informed into control centre, and corresponding disposal options are taken according to current break reason.The major-minor intelligent patrol detection equipment of the efficient Collaborative Control of the present invention, realize real-time and precise judgement, control to equipment running status: real-time perfoming monitoring, decision and action behavior realize that power grid is interactive, realize optimal control and strong self-healing.
Description
Technical field
The present invention relates to electric power inspection maintenance field more particularly to the fast linked places of failure of substation's power station multidimensional data
Set method.
Background technique
In the existing station course of work, since artificial intelligence direction was quickly grown in recent years, data reserves
With calculate equipment fast development so that we enter " big data era ".And in electric system D5000 and auxiliary control/
Temp measuring system carries out real time high-speed acquisition to full electric network key data, stores a large amount of grid operation data, " electric power for many years
System big data " direction is also come into being, and the current break reason to use machine learning algorithm to study electric system provides
Opportunity.
Summary of the invention
The technical problem to be solved by the invention is to provide a kind of failure of substation power station multidimensional data is fast linked
Method of disposal prevents maloperation and invalid operation to realize rapid identification reason and Disposal Measures in current break, improves inspection
Repair efficiency.
In order to solve the above technical problems, the present invention adopts the following technical scheme:
A kind of substation's multidimensional data operation interactive controlling platform, including multidimensional data operation interactive controlling platform host,
Crusing robot and control centre, the multidimensional data operation interactive controlling platform host and D5000 power scheduling monitoring are flat
Platform, electric power PMS system, electric power OMS system, on-line monitoring system, auxiliary control system and crusing robot connection are to acquire operation
It information and is controlled, when D5000 power scheduling monitor supervision platform monitors that current break occurs, multidimensional data operation interaction control
Platform host processed is analyzed and processed the operation information of acquisition, exports current break reason according to processing result, and according to electricity
Flow the Causes for Mutation control crusing robot work.
The present invention also provides a kind of fast linked methods of disposal of the failure based on substation's multidimensional data, using above-mentioned
Substation's multidimensional data operation interactive controlling platform carries out linkage disposition, includes the following steps:
S1 divides the operation information of acquisition when D5000 power scheduling monitor supervision platform monitors that current break occurs
Analysis processing, identify current break reason, wherein current break occurrence cause include system occur accident, normal service need and
Dispatching running way adjustment;
The current break reason short message of identification is informed control centre by S2, and is taken according to current break reason corresponding
Disposal options.
Optionally, when dimension data operation interactive controlling platform host identification current break occurrence cause is that accident occurs for system
When, using following disposal options:
(1) identification switch machinery indicating positions, short message mode notify setting personnel;
(2) identification protecting pressing plate state, short message mode notify setting personnel;
(3) identification protecting device indicator light, short message mode notify setting personnel;
Crusing robot performs the following operations:
(1) identification switch mechanical location;
(2) after switching mechanical location identification mission, crusing robot occurs according to D5000 power scheduling monitor supervision platform
Current break equipment situation carries out spy to relevant device and patrols;
(3) if there is manual control request in midway, work at present should be paused immediately, manual control is obeyed, to manual control knot
Shu Hou should continue current spy and patrol task;
(4) result will be maked an inspection tour and setting personnel is informed with short message mode.
Optionally, when dimension data operation interactive controlling platform host identification current break occurrence cause is normal service needs
When, using following disposal options:
(1) actual conditions, multidimensional data operation interaction control are mutated according to device current in D5000 power scheduling monitor supervision platform
Platform host processed accurate inspection range during confining maintenance automatically, and operation information in transfer order system is transferred in real time, determine inspection
Time started;
(2) it determines emphasis equipment routing inspection equipment: being determined according to robot history temperature measurement data, Yi Fare when determining high current
Equipment;Historic defects equipment is determined whether according to electric power PMS system defect record;The division of both the above equipment is attached most importance to and is patrolled
Equipment is examined, in later period O&M scheme, the emphasis equipment routing inspection frequency is higher than other equipment;
(3) D5000 power scheduling monitor supervision platform: machine utilization within the scope of the accurate inspection of real-time monitoring, when more than different limits
When volume, multidimensional data operation interactive controlling platform host provides corresponding alarm and inspection scheme automatically;
(4) crusing robot: repair apparatus carry out daily after using as a servant again primary spy patrol, infrared measurement of temperature;
(5) on-line monitoring system real-time monitoring ambient temperature and humidity;
(6) auxiliary control system: photographic technique device context situation, it is automatic to adjust according to the live data of the Temperature and Humidity module of acquisition
Save auxiliary control equipment in hyperbaric chamber, adjustment equipment running environment.
Optionally, when operation/maintenance data operation interactive controlling platform host identification current break occurrence cause is management and running side
When formula adjusts, using following disposal options:
(1) actual conditions, multidimensional data operation interaction control are mutated according to device current in D5000 power scheduling monitor supervision platform
Platform host processed accurate inspection range during confining maintenance automatically, and operation information in transfer order system is transferred in real time, determine inspection
Time started;
(2) machine utilization within the scope of the accurate inspection of D5000 power scheduling monitor supervision platform real-time monitoring, when more than different limits
When, multidimensional data operation interactive controlling platform host provides corresponding alarm and inspection scheme automatically;
(3) crusing robot identifies that carrying out the position of the switch to the switch of remote control operation determines by visible light, to precisely patrolling
Equipment timing is carried out spy and is patrolled in inspection range;
(4) on-line monitoring system real-time monitoring ambient temperature and humidity;
(5) auxiliary control system, photographic technique device context situation are automatic to adjust according to the live data of the Temperature and Humidity module of acquisition
Save auxiliary control equipment in hyperbaric chamber, adjustment equipment running environment.
Optionally, it includes: a. infrared measurement of temperature that the spy of crusing robot, which patrols content,;B. main transformer oil, oil level, SF6 pressure value,
Arrester Leakage Current;C. switch contact temperature, ambient temperature and humidity.
Optionally, the method that multidimensional data operation interactive controlling platform host is analyzed and processed the operation information of acquisition
Include the following steps:
Step (1) acquires the historical data under each various methods of operation in interval, is formed related with current break original
Data;
Step (2) extracts feature from initial data, and determines mutation label, utilizes D5000 power scheduling monitor supervision platform
Current capacity data determine whether that current break, which occurs, extracts all kinds of electrical measure features, as rear for all kinds of methods of operation
The feature input of continuous XGBoost algorithm, is consequently formed the sample data for modeling current break;
Step (3) carries out model training using the sample data of acquisition using XGBoost algorithm;
Step (4), after XGBoost model training is mature, the power grid recorded according to D5000 power scheduling monitor supervision platform
Real-time traffic information is configured to the electrical measure feature of reflection power network current mutation, inputs XGBoost model, can sentence in real time
Current break brought by the other certain possible failures of electric system.
Optionally, the minimum electricity for just acquiring scale when the electrical measure feature of reflection power network current mutation includes: current break
Flow valuve;D5000 power scheduling monitor supervision platform protection signal content when current break;D5000 power scheduling when current break
Monitor supervision platform switch trip signal content;D5000 power scheduling monitor supervision platform switchs remote control operation signal content when current break;
Substation topology figure is corresponded to when current break, if can unlink again with cyclization;Each equipment of crusing robot is anti-when current break
Feedback reading;On-line monitoring system is read when current break.
Optionally, in the training process, the feature different for Type Ⅰ Ⅱ error severity during current break prediction,
It introduces and notices that force coefficient is modified the loss function of XGBoost algorithm, so that prediction case of the model to unstable sample
It reduces.
Optionally, it is used for using logistic function by model output probability, for measuring the output of XGBoost model
The degree of reliability prevents part misprediction.
The present invention by adopting the above technical scheme, has the following beneficial effects:
1, information barrier between multisystem is got through, sufficiently integrates electric network information resource, building resource-sharing, system is steady, answers
With efficient O&M internet of things data application system: by real-time data capture, calculating in real time, real-time synchronization, relational data
Library, multidimensional data show, multidimensional data analysis engine (Mondrian), and the construction of buffer update sufficiently obtains number between each system
According to, using standard agreement, by rapid communication network, status information is uploaded to each control layer, in certain intelligent control layer,
According to the theory of optimal control and algorithm, control information and decision are formed, then feeds back to smart machine, completes device action.
2, the major-minor intelligent patrol detection equipment of efficient Collaborative Control realizes real-time and precise judgement, control to equipment running status:
Real-time perfoming monitoring, decision and action behavior realize that power grid is interactive, realize optimal control and strong self-healing.
3, it makes up artificial cycle and makes an inspection tour insufficient, saving human resources, improve fortune inspection efficiency, change manufacturing management system:
The task amount of manual inspection greatly reduces by the combination to robot system, on-line monitoring system, auxiliary control system, mitigates
The pressure of manual inspection improves the frequency of inspection and the efficiency and accuracy of fortune inspection.
The specific technical solution of the present invention and its advantages will carry out detailed in the following detailed description
Explanation.
Detailed description of the invention
Present invention will be further described below with reference to the accompanying drawings and specific embodiments:
Fig. 1 is XGboost algorithm flow chart.
Specific embodiment
Embodiment one
The present embodiment provides substation's multidimensional datas to run interactive controlling platform, including multidimensional data runs interactive controlling master
Machine, crusing robot, D5000 power scheduling monitor supervision platform, electric power PMS system, electric power OMS system, on-line monitoring system, auxiliary control
System.
The present invention can be achieved electric current reason and quickly position and automatic disposal function.Multidimensional data runs interactive controlling platform master
Machine is used to interact the relevant information of sequential operation in substation, is analyzed and processed to input information, logic judgment, according to processing
As a result current break reason is exported, realizes that the robot of specific programming operations is maked an inspection tour, equipment position is overhauled.
Multidimensional data operation interactive controlling platform host can be connect by serial ports with monitoring system, collection monitoring system letter
Breath, such as above-mentioned D5000 power scheduling monitor supervision platform, electric power PMS system, electric power OMS system, on-line monitoring system.
Multidimensional data, which runs interactive controlling platform host, to be connected by modem and control centre or/and centralized control center
It connects, realizes resource-sharing and information exchange.
Embodiment two
The present embodiment provides a kind of fast linked methods of disposal of the failure based on substation's multidimensional data, including following step
It is rapid:
S1 believes when monitoring that current break occurs in D5000 according to protection act information, switch changed position etc. in D5000
Breath analyzes accident, and confines accident range, and short message informs scheduling and production command centre, can recognize as current break generation
Reason is that accident occurs for system, disposal options:
(1) identification switch machinery indicating positions, short message mode notice rush for field personnel and related management personnel.
(2) identification protecting pressing plate state, short message mode notice rush for field personnel and related management personnel.Identification is protected
Protection unit indicator light, short message mode notice rush for field personnel and related management personnel.
Outdoor crusing robot:
(1) identification switch mechanical location.
(2) after switching mechanical location identification mission, robot system is right according to D5000 current break equipment situation
Relevant device carries out spy and patrols.(spy patrols content: a. infrared measurement of temperature;B. main transformer oil, oil level, SF6 pressure value, arrester leakage electricity
The reading of the visible lights such as stream is made a copy of;C. switch contact temperature is made a copy of in Shi Neigua rail robot identification;4. transferring corresponding history to make an inspection tour
Data (for 24 hours -15 days);5. ambient temperature and humidity is read).
(3) if midway has manual control to request (need to carry out supplement tour to faulty equipment), work at present should be paused immediately,
Obey manual control, to manual control after, current spy should be continued and patrol task.
(4) result will be maked an inspection tour and leader and operator on duty is informed with short message mode.
S2, when monitoring that current break occurs in D5000, platform stops labour application according to oms system, when extraction stops labour
Between, stop use as a servant equipment, judge risk class, be identified as current break reason normal service needs, disposal options:
(1) actual conditions are mutated according to device current in d5000, accurate inspection model during handing over control platform to confine maintenance automatically
Enclose [note: following patrol strategy is directed to the equipment confined in range].And operation information in transfer order system is transferred in real time, determination is patrolled
The inspection time started (all stops the labour operation end time).
(2) emphasis equipment routing inspection equipment is determined
2.1 determine according to robot history temperature measurement data, determine easy heat-producing device when high current.
2.2 determine whether historic defects equipment according to pms defect record.
2.3 divide both the above equipment for emphasis inspection device, in later period O&M scheme, the emphasis equipment routing inspection frequency
It is higher than other equipment.
(3) machine utilization within the scope of the accurate inspection of D5000 system real-time monitoring hands over control platform when being more than different limits
Automatically corresponding alarm and inspection scheme are provided.
(4) robot system
Repair apparatus carry out daily after using as a servant again primary spy patrol, infrared measurement of temperature;(specifically make an inspection tour content: a. infrared measurement of temperature;B. it leads
Become the reading of the visible lights such as oil temperature, oil level, SF6 pressure value, arrester Leakage Current to make a copy of;C. Shi Neigua rail robot identification is made a copy of
Switch contact temperature;It transfers corresponding history and makes an inspection tour data (for 24 hours -15 days)).
(5) ambient temperature and humidity is read
Real-time monitoring (main transformer) oil chromatography, core current, the micro- water of GIS, partial discharge;(route) leakage current of an arrester reading;
Transfer corresponding Historical Monitoring data (for 24 hours -15 days).
(6) auxiliary control system
Photographic technique device context situation;According to the live data of the Temperature and Humidity module of acquisition, high pressure Interior Space is automatically adjusted
The auxiliary controls equipment such as tune, blower, dehumidifier, adjustment equipment running environment.
S3, when monitoring that current break occurs in D5000, platform is according to the signal for having switch remote control operation in d5000
The mode of operation unlinked again with first cyclization can recognize as current break reason to be that dispatching running way adjusts, disposal options:
(1) it confines accurate inspection range: actual conditions is mutated according to device current in d5000, control platform is handed over to confine automatically
Accurate inspection range [following patrol strategy is directed to the equipment confined in range] during maintenance.
(2) D5000 system
Machine utilization within the scope of the accurate inspection of real-time monitoring hands over control platform to provide automatically accordingly when being more than different limits
Alarm and inspection scheme.
(3) robot system
3.1. robot identifies that carrying out the position of the switch to the switch of remote control operation determines by visible light.35kv is switched by room
The interior crusing robot on-site identification position of the switch.
Within the scope of 3.2 pairs of accurate inspections equipment 1h 8h for 24 hours 48h carry out primary spy patrol, infrared measurement of temperature.(in specific tour
Hold: a. infrared measurement of temperature;B. the visible lights such as main transformer oil, oil level, SF6 pressure value, arrester Leakage Current reading is made a copy of;C. in room
Switch contact temperature is made a copy of in the identification of Gua Gui robot;It transfers corresponding history and makes an inspection tour data (for 24 hours -15 days), ambient temperature and humidity reading
It takes).
(4) on-line monitoring system
Real-time monitoring (main transformer) oil chromatography, core current, the micro- water of GIS, partial discharge;(route) leakage current of an arrester reading;
Transfer corresponding Historical Monitoring data (for 24 hours -15 days).
(5) auxiliary control system
Photographic technique device context situation;According to the live data of the Temperature and Humidity module of acquisition, high pressure Interior Space is automatically adjusted
The auxiliary controls equipment such as tune, blower, dehumidifier, adjustment equipment running environment.
The present invention is according to power grid enterprises' characteristics of the production and operation, and by analyzing the practical business of power transformation, system occurs
Cause of accident is as follows: according to the information such as protection act information, switch changed position in D5000, analyzing accident, and confines accident range.
And short message informs scheduling and production command centre.Normal service needs: stopping labour application+each according to having in current break+oms system
Grade scheduling preliminary command has in relevant information+d5000 and has relevant device remote control operation information.Dispatching running way adjustment: current break+
There is switch remote control operation+elder generation's cyclization Open Loop Operation again in d5000.Equipment deficiency is abnormal: current break+double back interconnection both ends system
Unite equal paired running when wherein a current reduction, another increase, specify corresponding robot to make an inspection tour instruction and equipment position
Maintenance solution.
As an implementation, multidimensional data operation interactive controlling platform host analyzes the operation information of acquisition
The method of processing uses the current break reason method of discrimination based on XGBoost algorithm, is quickly distinguished with realizing in current break
Other reason and abort situation for deep maintenance and avoid failure from providing quick judgement tool.
Refering to what is shown in Fig. 1, XGBoost (ExtremeGradientBoosting) is a kind of by one group of Weak Classifier
Iterate to calculate the integrated study frame to realize Accurate classification effect.XGBoost can utilize the multi-threaded parallel meter of CPU automatically
It calculates, increases regular terms again in the realization of algorithm, greatly improve model calculation efficiency, and make the extensive energy of model
Power gets a promotion, and also has a clear superiority in Distributed Application.Compared to traditional GBDT
(GradientBoostDecisionTree), Taylor second order information has been used on XGBoost algorithm, has been used simultaneously in realization
Row/multicore calculation, training convergence more rapidly, accuracy rate it is higher.Currently, in shop sales forecast, event category and client
Behavior prediction etc., XGBoost have been used widely.
Data and feature determine the upper limit of machine learning, and model and algorithm only approach this upper limit, therefore
Feature Engineering is extremely important.The data that PMU is acquired in electric system are higher-dimension timing, and as electric system scale increases,
Data volume is also bigger, algorithm arithmetic speed is required higher, it is therefore necessary to initial data construction feature amount.Construction feature
Amount need to follow following principle:
1) characteristic quantity can be well reflected the generation of current break;
2) characteristic quantity is unrelated with electric system scale;
3) calculating of characteristic quantity should can guarantee temporal rapidity.
First according to the variation of operating status in electric system before and after current break, and selected with reference to other researchers in feature
It is as shown in table 1 to have chosen 6 features for the experience selected.
According to features described above, first with the correlation between Pearson correlation coefficient analysis feature, come judging characteristic it
Between whether there is information redundancy, related coefficient size indicates the linearly related degree between two variables fi and fj, calculates public
Formula is shown below:
In formula: the value range of c is -1≤c≤1.If usually | c | > 0.5, it is believed that the two has linear dependence;C=0
It represents completely not linearly related;C > 0, which is represented, to be positively correlated.
The result finally calculated according to aforesaid way retains and deletes corresponding feature progress aspect of model input data and builds
Mould.
Current break reason method of discrimination based on XGBoost algorithm, includes the following steps:
Step (1), using D5000, PMS, OMS, on-Line Monitor Device under the various methods of operation, the electricity that is respectively spaced
The data such as stream, load, temperature;The a large amount of and related initial data of current break is consequently formed.
Step (2) extracts feature from initial data, and determines mutation label;Utilize D5000 system power load data
Determine whether that current break occurs;For all kinds of methods of operation, all kinds of electrical measure features are extracted, are calculated as subsequent XGBoost
The feature of method inputs;It is consequently formed a certain number of for modeling the sample data of current break.
Step (3) using XGBoost algorithm and carries out applicability improvement, carries out model instruction using the sample data of acquisition
Practice;In the training process, the feature different for Type Ⅰ Ⅱ error severity during current break prediction, introduces attention system
The loss function of several pairs of XGBoost algorithms is modified, so that model reduces the prediction case of unstable sample;It uses
Logistic function is used for model output probability, and for measuring the degree of reliability of XGBoost model output, prevention part is accidentally
Prediction.
Step (4) after XGBoost model training is mature, is believed according to the power grid real time execution that D5000 system is recorded
Breath is configured to the electrical measure feature of reflection power network current mutation;XGBoost model is inputted, it can real time discriminating electric system
Current break brought by certain possible failures.
Wherein, feature refers to station equipment combined electric amount, and need to extract is history mutation time of origin every interior master
Change, route, switch electric current value tag, the numerical value set combined by these features is called mutation label.
Be able to reflect power network current mutation electrical measure feature include history mutation occur when interval in main transformer, line
Road, switch electric current value tag, include each high, normal, basic main transformer current value, line electricity flow valuve, switching current value, when history is normal
When interval in main transformer, route, switch electric current value tag, include each high, normal, basic main transformer current value, line electricity flow valuve, switching current
Value.
The position of current break is determined, when current break occurs, specifically with equipment in a substation interval
Current value variation stability bandwidth come determine whether occur current break, such as: the 1# main transformer current value amplitude and history of Ke Yanbian
The same period is not inconsistent, or deviates practical experience amplitude, then knows mutation hair on the 1# main transformer of Ke Yanbian.
The above-mentioned current break reason method of discrimination based on XGBoost algorithm, has the advantages that
(1) since XGBoost can use multithreads computing automatically, arithmetic speed is fast, is suitble to the extensive electricity of processing
Force data.
(2) in model on increase the regularization term constituted with the complexity of tree so that its generalization ability gets a promotion, have
Effect solves over-fitting;
(3) it since XGBoost is Tree-structure Model, does not need to various data normalizations collected in electric system,
And missing values caused by capable of being effectively treated for some reason, it is suitable for the current break of spacing current.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, is familiar with
The those skilled in the art should be understood that the present invention includes but is not limited to content described in specific embodiment above.It is any
Modification without departing from function and structure principle of the invention is intended to be included in the range of claims.
Claims (10)
1. substation's multidimensional data runs interactive controlling platform, it is characterised in that: run interactive controlling platform including multidimensional data
Host, crusing robot and control centre, the multidimensional data operation interactive controlling platform host and D5000 power scheduling are supervised
Platform, electric power PMS system, electric power OMS system, on-line monitoring system, auxiliary control system and crusing robot connection are controlled to acquire
It operation information and is controlled, when D5000 power scheduling monitor supervision platform monitors that current break occurs, multidimensional data operation is handed over
Mutually control platform host is analyzed and processed the operation information of acquisition, exports current break reason, and root according to processing result
According to the control crusing robot work of current break reason.
2. the fast linked method of disposal of failure based on substation's multidimensional data, using substation's multidimensional described in claim 1
Data run interactive controlling platform carries out linkage disposition, it is characterised in that includes the following steps:
S1 carries out at analysis the operation information of acquisition when D5000 power scheduling monitor supervision platform monitors that current break occurs
Reason identifies current break reason, and wherein current break occurrence cause includes that accident occurs for system, normal service needs and scheduling
Method of operation adjustment;
The current break reason short message of identification is informed control centre, and takes corresponding disposition according to current break reason by S2
Mode.
3. the fast linked method of disposal of the failure according to claim 2 based on substation's multidimensional data, it is characterised in that:
When dimension data operation interactive controlling platform host identification current break occurrence cause is that accident occurs for system, using following disposition
Mode:
(1) identification switch machinery indicating positions, short message mode notify setting personnel;
(2) identification protecting pressing plate state, short message mode notify setting personnel;
(3) identification protecting device indicator light, short message mode notify setting personnel;
Crusing robot performs the following operations:
(1) identification switch mechanical location;
(2) after switching mechanical location identification mission, according to D5000 power scheduling monitor supervision platform electric current occurs for crusing robot
It is mutated equipment situation, spy is carried out to relevant device and is patrolled;
(3) if there is manual control request in midway, work at present should be paused immediately, obey manual control, to manual control after,
It should continue current spy and patrol task;
(4) result will be maked an inspection tour and setting personnel is informed with short message mode.
4. the fast linked method of disposal of the failure according to claim 2 based on substation's multidimensional data, it is characterised in that:
When dimension data operation interactive controlling platform host identification current break occurrence cause needs for normal service, using following disposition
Mode:
(1) actual conditions are mutated according to device current in D5000 power scheduling monitor supervision platform, it is flat that multidimensional data runs interactive controlling
Platform host accurate inspection range during confining maintenance automatically, and operation information in transfer order system is transferred in real time, determine that inspection starts
Time;
(2) it determines emphasis equipment routing inspection equipment: being determined according to robot history temperature measurement data, determine that easily fever is set when high current
It is standby;Historic defects equipment is determined whether according to electric power PMS system defect record;Both the above equipment is divided as emphasis inspection
Equipment, in later period O&M scheme, the emphasis equipment routing inspection frequency is higher than other equipment;
(3) D5000 power scheduling monitor supervision platform: machine utilization within the scope of the accurate inspection of real-time monitoring, when being more than different limits,
Multidimensional data operation interactive controlling platform host provides corresponding alarm and inspection scheme automatically;
(4) crusing robot: repair apparatus carry out daily after using as a servant again primary spy patrol, infrared measurement of temperature;
(5) on-line monitoring system real-time monitoring ambient temperature and humidity;
(6) auxiliary control system: photographic technique device context situation, according to the live data of the Temperature and Humidity module of acquisition, automatic adjustment is high
Auxiliary control equipment in pressure chamber, adjustment equipment running environment.
5. the fast linked method of disposal of the failure according to claim 2 based on substation's multidimensional data, it is characterised in that:
When operation/maintenance data operation interactive controlling platform host identification current break occurrence cause is that dispatching running way adjusts, using such as
Lower disposal options:
(1) actual conditions are mutated according to device current in D5000 power scheduling monitor supervision platform, it is flat that multidimensional data runs interactive controlling
Platform host accurate inspection range during confining maintenance automatically, and operation information in transfer order system is transferred in real time, determine that inspection starts
Time;
(2) machine utilization within the scope of the accurate inspection of D5000 power scheduling monitor supervision platform real-time monitoring, when being more than different limits,
Multidimensional data operation interactive controlling platform host provides corresponding alarm and inspection scheme automatically;
(3) crusing robot identifies that carrying out the position of the switch to the switch of remote control operation determines by visible light, to accurate inspection model
Interior equipment timing development spy is enclosed to patrol;
(4) on-line monitoring system real-time monitoring ambient temperature and humidity;
(5) auxiliary control system, photographic technique device context situation, according to the live data of the Temperature and Humidity module of acquisition, automatic adjustment is high
Auxiliary control equipment in pressure chamber, adjustment equipment running environment.
6. the fast linked method of disposal of failure based on substation's multidimensional data according to claim 3 or 4 or 5,
It is characterized by: it includes: a. infrared measurement of temperature that the spy of crusing robot, which patrols content,;B. main transformer oil, oil level, SF6 pressure value, lightning-arrest
Device Leakage Current;C. switch contact temperature, ambient temperature and humidity.
7. the fast linked method of disposal of the failure according to claim 2 based on substation's multidimensional data, it is characterised in that:
The method that multidimensional data operation interactive controlling platform host is analyzed and processed the operation information of acquisition includes the following steps:
Step (1) acquires the historical data under each various methods of operation in interval, forms original number related with current break
According to;
Step (2) extracts feature from initial data, and determines mutation label, utilizes D5000 power scheduling monitor supervision platform electric current
Load data determines whether that current break, which occurs, extracts all kinds of electrical measure features, as subsequent for all kinds of methods of operation
The feature of XGBoost algorithm inputs, and the sample data for modeling current break is consequently formed;
Step (3) carries out model training using the sample data of acquisition using XGBoost algorithm;
Step (4), after XGBoost model training is mature, the power grid recorded according to D5000 power scheduling monitor supervision platform is real-time
Operation information is configured to the electrical measure feature of reflection power network current mutation, inputs XGBoost model, can real time discriminating electricity
Current break brought by the certain possible failures of Force system.
8. the fast linked method of disposal of the failure according to claim 7 based on substation's multidimensional data, it is characterised in that:
The electrical measure feature for reflecting power network current mutation minimum current value for just acquiring scale when including: current break;When current break
D5000 power scheduling monitor supervision platform protection signal content;D5000 power scheduling monitor supervision platform switch trip when current break
Signal content;D5000 power scheduling monitor supervision platform switchs remote control operation signal content when current break;To strain when current break
Power station topological diagram, if can be unlinked again with cyclization;Each equipment feedback readings of crusing robot when current break;When current break
On-line monitoring system reading.
9. the fast linked method of disposal of the failure according to claim 7 based on substation's multidimensional data, it is characterised in that:
In the training process, the feature different for Type Ⅰ Ⅱ error severity during current break prediction, introduces and pays attention to force coefficient
The loss function of XGBoost algorithm is modified, so that model reduces the prediction case of unstable sample.
10. the fast linked method of disposal of the failure according to claim 7 based on substation's multidimensional data, feature exist
In: it is used for using logistic function by model output probability, for measuring the degree of reliability of XGBoost model output.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910452000.8A CN110349048B (en) | 2019-05-28 | 2019-05-28 | Substation multidimensional data operation interactive control platform and fault handling method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910452000.8A CN110349048B (en) | 2019-05-28 | 2019-05-28 | Substation multidimensional data operation interactive control platform and fault handling method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110349048A true CN110349048A (en) | 2019-10-18 |
CN110349048B CN110349048B (en) | 2023-05-26 |
Family
ID=68174012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910452000.8A Active CN110349048B (en) | 2019-05-28 | 2019-05-28 | Substation multidimensional data operation interactive control platform and fault handling method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110349048B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111555162A (en) * | 2020-04-06 | 2020-08-18 | 连云港市港圣开关制造有限公司 | High-efficient intelligent power distribution operation and maintenance system |
CN116054417A (en) * | 2023-03-16 | 2023-05-02 | 山东联行智能科技有限公司 | Monitoring system and method for transformer substation |
CN117034174A (en) * | 2023-09-26 | 2023-11-10 | 国网安徽省电力有限公司经济技术研究院 | Transformer substation equipment abnormality detection method and system |
CN117081249A (en) * | 2023-08-17 | 2023-11-17 | 广东正力通用电气有限公司 | Automatic monitoring management platform for power supply line and line fault identification method |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4689752A (en) * | 1983-04-13 | 1987-08-25 | Niagara Mohawk Power Corporation | System and apparatus for monitoring and control of a bulk electric power delivery system |
JPH04185206A (en) * | 1990-11-19 | 1992-07-02 | Toshiba Corp | Controller for electric vehicle driving motor |
US20120212862A1 (en) * | 2009-08-20 | 2012-08-23 | State Grid Corporation Of China | Method and device for limiting secondary arc current of extra-high voltage/ultra-high voltage double circuit lines on the same tower |
CN103117602A (en) * | 2013-03-11 | 2013-05-22 | 辽宁省电力有限公司 | Method for automatically reporting composite accidents in relay protection intelligence online running management system |
CN104052151A (en) * | 2014-05-14 | 2014-09-17 | 国家电网公司 | Intelligent auxiliary comprehensive monitoring and D5000 comprehensive linkage platform of transformer substation |
CN106685704A (en) * | 2016-12-13 | 2017-05-17 | 云南电网有限责任公司电力科学研究院 | Checking method and checking device for secondary equipment failure of transformer substations |
JP2017130391A (en) * | 2016-01-21 | 2017-07-27 | 国立大学法人東京工業大学 | Circuit breaker |
CN108551167A (en) * | 2018-04-25 | 2018-09-18 | 浙江大学 | A kind of electric power system transient stability method of discrimination based on XGBoost algorithms |
CN109101906A (en) * | 2018-07-27 | 2018-12-28 | 中国南方电网有限责任公司超高压输电公司贵阳局 | A kind of converting station electric power equipment infrared image exception real-time detection method and device |
-
2019
- 2019-05-28 CN CN201910452000.8A patent/CN110349048B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4689752A (en) * | 1983-04-13 | 1987-08-25 | Niagara Mohawk Power Corporation | System and apparatus for monitoring and control of a bulk electric power delivery system |
JPH04185206A (en) * | 1990-11-19 | 1992-07-02 | Toshiba Corp | Controller for electric vehicle driving motor |
US20120212862A1 (en) * | 2009-08-20 | 2012-08-23 | State Grid Corporation Of China | Method and device for limiting secondary arc current of extra-high voltage/ultra-high voltage double circuit lines on the same tower |
CN103117602A (en) * | 2013-03-11 | 2013-05-22 | 辽宁省电力有限公司 | Method for automatically reporting composite accidents in relay protection intelligence online running management system |
CN104052151A (en) * | 2014-05-14 | 2014-09-17 | 国家电网公司 | Intelligent auxiliary comprehensive monitoring and D5000 comprehensive linkage platform of transformer substation |
JP2017130391A (en) * | 2016-01-21 | 2017-07-27 | 国立大学法人東京工業大学 | Circuit breaker |
CN106685704A (en) * | 2016-12-13 | 2017-05-17 | 云南电网有限责任公司电力科学研究院 | Checking method and checking device for secondary equipment failure of transformer substations |
CN108551167A (en) * | 2018-04-25 | 2018-09-18 | 浙江大学 | A kind of electric power system transient stability method of discrimination based on XGBoost algorithms |
CN109101906A (en) * | 2018-07-27 | 2018-12-28 | 中国南方电网有限责任公司超高压输电公司贵阳局 | A kind of converting station electric power equipment infrared image exception real-time detection method and device |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111555162A (en) * | 2020-04-06 | 2020-08-18 | 连云港市港圣开关制造有限公司 | High-efficient intelligent power distribution operation and maintenance system |
CN116054417A (en) * | 2023-03-16 | 2023-05-02 | 山东联行智能科技有限公司 | Monitoring system and method for transformer substation |
CN116054417B (en) * | 2023-03-16 | 2023-09-19 | 山东联行智能科技有限公司 | Monitoring system and method for transformer substation |
CN117081249A (en) * | 2023-08-17 | 2023-11-17 | 广东正力通用电气有限公司 | Automatic monitoring management platform for power supply line and line fault identification method |
CN117034174A (en) * | 2023-09-26 | 2023-11-10 | 国网安徽省电力有限公司经济技术研究院 | Transformer substation equipment abnormality detection method and system |
CN117034174B (en) * | 2023-09-26 | 2023-12-29 | 国网安徽省电力有限公司经济技术研究院 | Transformer substation equipment abnormality detection method and system |
Also Published As
Publication number | Publication date |
---|---|
CN110349048B (en) | 2023-05-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110349048A (en) | Substation's multidimensional data runs interactive controlling platform and failure method of disposal | |
CN108336725B (en) | Power grid dispatching equipment monitoring management and intelligent analysis system | |
CN112788142B (en) | Intelligent edge Internet of things gateway supporting multi-sensor access | |
CN207965544U (en) | A kind of smart substation equipment fault pre-alarming mechanism | |
CN111474444B (en) | Line fault power restoration decision method based on knowledge graph | |
CN113466597A (en) | Intelligent detection method for equipment state of rail transit power supply system | |
CN107124291A (en) | A kind of adjusting device monitoring analysis system and method based on big data | |
CN106997513A (en) | The distribution net equipment state evaluation system analyzed based on big data | |
CN105978145B (en) | Secondary system of intelligent substation information fusion system | |
CN110245846A (en) | Current break reason method of discrimination based on XGBoost algorithm | |
Ma et al. | Research Progress of Distribution Network Digitization Based on Digital Twin Technology | |
Xu et al. | Research on fault diagnosis method of power grid based on artificial intelligence | |
CN116026410A (en) | Automatic inspection system based on computer intelligent technology | |
CN109217311B (en) | Power distribution network operation state control and evaluation method | |
CN113726007B (en) | Intelligent recorder master station management method and system based on transparent system architecture | |
CN111539642B (en) | Object-oriented power plant data acquisition and processing system and method thereof | |
CN205081568U (en) | High performance transformer intelligent control ware | |
CN113689064A (en) | Transparent transformer substation management system architecture | |
CN202888970U (en) | Transformer substation secondary equipment data and state on-line collecting device | |
CN111401481A (en) | Power grid operation mode automatic identification method based on multi-source information fusion technology | |
CN113541320A (en) | Power grid power failure and restoration visual monitoring method and system and storage medium | |
Wang et al. | Application of lambda architecture in intelligent operation and maintenance system of rail transit vehicles | |
Min | Improved Fault Prediction Algorithm of High-Speed EMUs based on PHM Technology | |
CN113269435B (en) | New energy station running state coupling monitoring and evaluating system | |
Liangzhi et al. | Research on fault prediction and diagnosis of power equipment based on big data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |