CN104239437A - Power-network-dispatching-oriented intelligent warning analysis method - Google Patents
Power-network-dispatching-oriented intelligent warning analysis method Download PDFInfo
- Publication number
- CN104239437A CN104239437A CN201410433102.2A CN201410433102A CN104239437A CN 104239437 A CN104239437 A CN 104239437A CN 201410433102 A CN201410433102 A CN 201410433102A CN 104239437 A CN104239437 A CN 104239437A
- Authority
- CN
- China
- Prior art keywords
- warning information
- alarm
- denoising
- frequent item
- warning
- 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
- 238000004458 analytical method Methods 0.000 title abstract description 5
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000005065 mining Methods 0.000 claims abstract description 24
- 238000003066 decision tree Methods 0.000 claims description 34
- 238000013459 approach Methods 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 17
- 239000000284 extract Substances 0.000 claims description 14
- 238000005259 measurement Methods 0.000 claims description 11
- 238000009412 basement excavation Methods 0.000 claims description 9
- 230000007423 decrease Effects 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000012217 deletion Methods 0.000 claims description 3
- 230000037430 deletion Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000001174 ascending effect Effects 0.000 claims description 2
- 230000003247 decreasing effect Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 abstract description 4
- 230000006698 induction Effects 0.000 abstract 1
- 230000000875 corresponding effect Effects 0.000 description 23
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 230000003203 everyday effect Effects 0.000 description 5
- 238000007418 data mining Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000033228 biological regulation Effects 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 230000015654 memory Effects 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- 230000011664 signaling Effects 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 229940074869 marquis Drugs 0.000 description 1
- VBUNOIXRZNJNAD-UHFFFAOYSA-N ponazuril Chemical compound CC1=CC(N2C(N(C)C(=O)NC2=O)=O)=CC=C1OC1=CC=C(S(=O)(=O)C(F)(F)F)C=C1 VBUNOIXRZNJNAD-UHFFFAOYSA-N 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
-
- 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
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Alarm Systems (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a power-network-dispatching-oriented intelligent warning analysis method. The method comprises the following steps that warning information of a power network is extracted and is preprocessed; denoising is carried out on the preprocessed warning information, and a denoising classification rule is obtained; the denoised warning information is subjected to induction and dupliction removal, warning combinations are generated, and the warning type to which each warning combination belongs is judged; each warning combination and the corresponding warning type are subjected to association rule mining, and a power network warning inference rule is obtained. After the intelligent warning analysis method is utilized, the filtering rate of noise data in the warning information and the comprehensiveness of the warning types are improved.
Description
Technical field
The present invention relates to a kind of alert analysis method, particularly relate to a kind of intelligent alarm analytical approach towards dispatching of power netwoks, belong to electric power system dispatching technical field.
Background technology
Along with the continuous expansion of electrical network scale, the warning information of generation is also growing, have every day several thousand even up to ten thousand warning information pour in control center.There is abnormal and necessary overhaul of the equipments will produce noise data because measuring equipment is in operation, cause the accuracy rate of warning information inadequate.Meanwhile, current alarm information directly presents to dispatcher at a terrific speed, does not form the electrical network inference rule with incidence relation, causes dispatcher cannot judge alarm type rapidly.
In order to ensure that warning information is accurate, and alarm type can be judged fast, need the feature in conjunction with electrical network self and the development of data mining technology in electrical network, usage data mining algorithm excavates mass alarm information, improve the noise filtering rate of warning information, excavate electrical network alarm inference rule.
At present, remove in noise and refinement inference rule at warning information both at home and abroad and adopt multiple method to study, achieve certain achievement.Specific as follows:
1) the intelligent alarm processor of rule-based formula is proposed.
2) method of artificial neural network is applied in intelligent warning system, is used for analyzing compound alarm.
3) sequential mode mining is used in warning information process, excavates warning information and successively report the incidence relation sent out in time.
4) decision tree is used in warning information denoising.
5) utilize the incidence relation between rough set acquisition warning information, obtain electrical network alarm regulation.
In Master's thesis " applied research of data mining technology in SCADA warning information the is analyzed " (North China Electric Power University that Pan Li delivers, 2006) in, propose and utilize decision Tree algorithms to carry out denoising to the warning information in SCADA, effectively can remove the noise data in warning information.Enter Chao, the paper " the electrical network alarm regulation based on rough set theory automatically extracts and applies " of the refined co-present of Liu Wenying, Liu Yongzhi and Zhao Lian (is published in " protecting electrical power system and control "; 08 phase in 2011) in; proposition utilizes rough set theory to analyze warning information, automatically extracts electrical network alarm regulation.
But, adopt decision Tree algorithms to carry out denoising in prior art, adopt rough set method to carry out network planning reasoning, still have the following disadvantages:
1) only utilize historical events table during denoising, not in conjunction with other relevant informations, cause the denoising time long.
2) decision Tree algorithms is not improved during denoising, cause classification accuracy to decline.
3) when utilizing coarse central algorithm to carry out rule-based reasoning, alarm type is only set as common alarm type, causes electrical network alarm inference rule comprehensive not.
When excavating mass alarm information, effectively can not improve the noise filtering rate of warning information, excavating comprehensive electrical network alarm inference rule, effectively can not meet the demand of electric system.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of intelligent alarm analytical approach towards dispatching of power netwoks.
For realizing above-mentioned goal of the invention, the present invention adopts following technical scheme:
Towards an intelligent alarm analytical approach for dispatching of power netwoks, comprise the steps:
Extract the warning information of electrical network, pre-service is carried out to it;
Carry out denoising to through pretreated warning information, and obtain denoising classifying rules;
Warning information after denoising is concluded, duplicate removal, generate alarm combination, for the alarm type that each alarm combination judges belonging to it;
The alarm combination alarm type corresponding with it is carried out to the excavation of correlation rule, obtain electrical network alarm inference rule.
Wherein more preferably, when carrying out pre-service to warning information, first warning information extracted, change, load process, remove incoherent warning information, be then associated with the record of examination in dispatching management information system by warning information, the threshold value according to setting removes obvious noise data, warning information is associated with the remote measurement value in SCADA, according to the change information of remote measurement value, judge electrical network warning information whether mistake, remove the warning information of mistake.
Wherein more preferably, when carrying out denoising to warning information, analyze through pretreated warning information, process, being formed to meet adopts the decision Tree algorithms of band Bayes node to carry out the data processing table of classifying, therefrom a part of data of random extraction are carried out classification and are formed denoising decision tree, judge whether warning information is noise data, and denoising decision tree is changed into denoising classifying rules.
Wherein more preferably, after denoising decision tree generates, use remaining data to infer as test data, judge that it belongs to normal data or noise data, and itself and the classification belonging to original are compared, assess the accuracy rate that it detects.
Wherein more preferably, when carrying out the excavation of correlation rule to the alarm combination Alarm Classification corresponding with it, first attribute corresponding with warning information for alarm combination is combined, generate new alarm combination, again new alarm combination is combined with corresponding alarm type, form alarm affairs, the warning information transaction table of multiple alarm affairs formation is carried out to the excavation of frequent item set, association rule mining is carried out to the frequent item set meeting minimum support threshold values, obtains electrical network alarm inference rule.
Wherein more preferably, use and improve FP growth algorithm carries out frequent item set excavation to the warning information transaction table that multiple alarm affairs are formed, comprise the steps:
Step 21, the warning information transaction table in scan database, finds out the set of candidate, and obtains their support counting; Successively decrease according to support counting and arrange the every of candidate, obtain gathering F, support in set F is less than the entry deletion of minimum support threshold values, obtains the set L of frequent item set;
Step 22, warning information transaction table again in scan database, item support being less than minimum support threshold values is deleted from each affairs, and successively decreasing according to every support counting rearranges the item in each affairs, obtains the warning information transaction table after processing;
Step 23, according to every support counting in set L, ascendingly constructs every database subset successively, and utilizes FP growth algorithm to carry out constrained frequent item sets mining to it;
Step 24, after the constraint frequent item collection of items all in set L excavates out successively, merges these constraint frequent item collection, obtains all frequent item sets of warning information transaction table.
Wherein more preferably, the ascending step constructing every database subset is successively as follows:
Step 231, the warning information transaction table in scan database after process, therefrom extracts all containing item I
iaffairs, delete support in these affairs and be less than the item of this support, obtain an I
idatabase subset;
Step 232, to database subset, utilizes FP growth algorithm to carry out comprising an I
iconstraint frequent item set mining.
Wherein more preferably, the constraint frequent item set mining utilizing FP growth algorithm to carry out comprising item comprises the steps:
Step 2321, utilizes database subset, and structure FP tree, and creates item head table, and that last in item head table indicates is an I
isupport counting and node chain information;
Step 2322, by last information indicated in item head table, constructs this conditional pattern base, constructs its condition FP and sets, and this condition FP tree is excavated the constraint frequent item collection comprising this, completes the constraint frequent item set mining on database subset.
Wherein more preferably, association rule mining is carried out to the frequent item set meeting minimum support threshold values, obtain electrical network alarm inference rule and comprise the steps:
Computing formula according to degree of confidence:
Obtain correlation rule:
(1) for each frequent item set l, all nonvoid subsets of l are produced;
(2) for each nonvoid subset s of l, if
then export rule
Alarm combination in frequent item set is set to A, and alarm classification is set to B, and min_conf is set to 100%, derives electrical network alarm inference rule be by above-mentioned correlation rule:
Intelligent alarm analytical approach provided by the present invention, by by the record of examination in warning information and OMS, and the remote measurement value in SCADA combines, conclude the decision attribute made new advances, reduce the denoising time, and improve denoising accuracy rate, the decision Tree algorithms of band Bayes node is used to carry out denoising to warning information, the defect that NB Algorithm cannot extract rule can be solved, can decision Tree algorithms be improved again, improve the accuracy rate that warning information carries out classifying.In addition, the present invention improves FP growth algorithm by use and produces frequent item set, and recycling correlation rule produces electrical network alarm inference rule, improves the efficiency of electrical network alarm inference rule generation and the comprehensive of alarm type.
Accompanying drawing explanation
Fig. 1 is in one embodiment of the present of invention, warning information is carried out to the process flow diagram of intelligent alarm analysis;
Fig. 2 is in one embodiment of the present of invention, carries out ETL processing procedure schematic diagram to warning information;
Fig. 3 is in one embodiment of the present of invention, the Star Model schematic diagram of alarm information record;
Fig. 4 is in one embodiment of the present of invention, warning information is carried out to the process flow diagram of denoising;
Fig. 5 is the process flow diagram using Bayes principle structure decision tree;
Fig. 6 is in one embodiment of the present of invention, the denoising decision-tree model schematic diagram of generation;
Fig. 7 is in one embodiment of the present of invention, the process flow diagram that alarm inference rule obtains.
Embodiment
Below in conjunction with the drawings and specific embodiments, technology contents of the present invention is described in further detail.
As shown in Figure 1, intelligent alarm analytical approach towards dispatching of power netwoks provided by the invention, comprise the steps: the warning information extracting electrical network, pre-service is carried out to the warning information of electrical network, statistics warning information every day and report monthly send out the relation of number of times with noise data, according to the ruuning situation of electrical network and the suitable threshold value of the requirements set of alarm degree of accuracy, when warning information every day and report monthly send out number of times exceed the threshold value of setting time, this warning information is evident as noise data, removes obvious noise data; Warning information is associated with remote measurement value, removes the warning information of mistake; Then adopt the decision Tree algorithms of band Bayes node to classify to warning information, judge whether warning information is noise data, and obtain denoising classifying rules; To the warning information after denoising, adopt and improve the excavation that FP growth algorithm carries out correlation rule, obtain the incidence relation between warning information and alarm type, obtain comprehensive electrical network alarm inference rule.Detailed specific description is done to this process below.
As shown in Figure 1, when pre-service is carried out to electrical network warning information, first ETL process (namely Extract-Transform-Load carries out the process of data pick-up, conversion, loading to warning information) is carried out to warning information.Electric network synthesis information data are many, and comprise noise data, deficiency of data, even have inconsistent data, carry out changing the quality that can improve data mining object, improve the accuracy of electrical network alarm inference rule to these data.
In one embodiment of the invention, warning information in item history lists is associated with the remote measurement value in the record of examination in OMS (dispatching management information system) and SCADA (data acquisition and supervisor control), noise data, deficiency of data and inconsistent data are tentatively rejected, as shown in Figure 2, required data are extracted from historical events table, SCADA and OMS, form alarm information record table after preliminary rejecting is carried out to noise data, deficiency of data and inconsistent data, leave in historical data base as data source.When warning information carries out ETL process, from historical data base, extract related data.In one embodiment of the invention, historical data after processing is left in historical data base with Star Model, set up the Star Model of warning information, as shown in Figure 3, the information of carrying out analyzing and processing and needing to be correlated with can be found fast according to the hub-and-spoke configuration of warning information.
After ETL process is carried out to warning information, remove obvious noise data, mainly leave out the warning information (i.e. incoherent warning information) corresponding to attribute that in alarm information record table, yardman does not pay close attention to.In one embodiment of the invention, the attribute should paid close attention to according to the requirements set yardman of intelligent grid supporting system technology forms required attribute.In addition, according to the relation of adding up the DAY_CNT report of every day (warning information send out number of times) and MONTH_CNT (warning information report monthly sends out number of times) and the noise data obtained, according to the ruuning situation of electrical network and the suitable threshold value of the requirements set of alarm degree of accuracy, when warning information every day and report monthly send out number of times exceed the threshold value of setting time, this warning information is evident as noise data, removes obvious noise data.
When power grid accident occurs time, one intuitively phenomenon be exactly scheduling picture on see on the electrical equipment relevant to trip breaker the remote measurement value such as voltage, electric current, power zero.Therefore, at the pretreatment stage of warning information, the change information of remote measurement value, be used as the subsidiary discriminant means whether wrong to electrical network warning information, realize the subsidiary discriminant to electrical network warning information with this, remove the warning information of mistake.
As shown in Figure 4, when the alarm information record table in historical data base carries out information pre-processing, after removing obvious noise data, extract required attribute and form data prediction table.Attribute in data prediction table comprises: ID, MSG_TYPE, ALMMSG, FAULT_INFO, SOUID, TIME and record of examination.As shown in table 1.
| ID | MSG_TYPE | ALMMSG | FAULT_INFO | SOUID | TIME | Record of examination |
Table 1 data prediction table
Wherein, the implication of the attribute comprised in data prediction table is as follows:
(1) ALMMSG: represent the concrete text description to alarm event.
(2) SOUID: represent the device numbering that alarm event occurs, unique device numbering that each equipment has it corresponding.
(3) MSG_TYPE: the type representing alarm event.
(4) FAULT_INFO: represent the current failure message of alarm event, as: for switch, to be divided into point, conjunction two kinds; For alarm, be divided into alarm, involution two kinds.
(5) TIME: represent that the report of alarm event sends out the moment.
(6) record of examination: represent that the moment occurs alarm event, whether this equipment is in maintenance.
After data prediction table generates, processing rule according to preserving in advance is in a database analyzed warning information, summarize the base attribute involved by warning information analyzed, calculate base attribute value, the base attribute table formed, the attribute in base attribute table comprises: ID, MSG_TYPE, DURATION, INTERVAL, RECORD, OCCURRENCE.As shown in table 2.
| ID | MSG_TYPE | DURATION | INTERVAL | RECORD | OCCURRENCE |
Table 2 base attribute table
Wherein, the implication of the attribute comprised in base attribute table is as follows:
(1) MSG_TYPE: directly take from data prediction table, defines identical with belonging to originally property.
(2) INTERVAL (report sends out interval): be used for representing that this report of certain abnormality alarming information sends out the time interval apart from last time, report was sent out.It is far-ranging discrete value.It is too short that report sends out interval, illustrates that the data that remote signalling is measured are unstable, can judge that warning information is noise data.
(3) DURATION (duration): be used for representing duration of warning information; i.e. time interval of sending out of alarm clearing and alarm report; when alarm occurs; corresponding protection act can be taked; for the involution of alarm, DURATION is the time interval alarm being dealt into involution from alarm report.It is far-ranging discrete value.Duration is too short, illustrates that the data that remote signalling is measured are unstable, can judge that warning information is noise data.
(4) OCCURRENCE (the previous day whether report send out): be used for representing whether warning information is reported the previous day and sent out.In electrical network, switch failure tripping operation, this kind of warning information of protection act seldom can be sent out with Times for continuous two days, when the alarm of these types report continuously send out time, can judge that this warning information is noise data rapidly
(5) RECORD (record of examination): be used for representing whether current alarm event is caused by normal service.
The property value of base attribute is analyzed, concludes new codomain, and Binding number Data preprocess table determine to analyze warning information needed for final attribute, the property value extracting respective attributes forms data processing table.Territory in data processing table comprises: ID, MSG_TYPE, DURATION, INTERVAL, OCCURRENCE, RECORD, DECISION.As shown in table 3.
| ID | MSG_TYPE | DURATION | INTERVAL | OCCURRENCE | RECORD | DECISION |
Table 3 data processing table
Wherein, the implication of the field comprised in data processing table is as follows:
(1) MSG_TYPE (alarm type): the type being used for representing current alarm information, in one embodiment of the invention, is divided into guard signal, anticipating signal, switching signal and alarm signal.
(2) RECORD (record of examination): be used for representing whether current alarm event is caused by normal service.In one embodiment of the invention, in order to effectively classify, be set with, without two attributes.If property value, for having, is directly judged to be noise data.
(3) INTERVAL (report send out an interval): be used for the time interval representing that warning information report is sent out, report corresponding to base attribute sends out interval, in order to effectively classify, it is risen to higher concept hierarchy, in one embodiment of the invention, long, medium and short three property values are set.It is short that a report interval is less than 1 hour corresponding property value, is more than or equal to 1 hour and is less than or equal in 6 hours corresponding property values, being greater than 6 hours corresponding property values long.
(4) DURATION (duration): the duration of this exception, corresponding to the duration of base attribute, its property value is also far-ranging discrete value, in order to effectively classify, it is risen to higher concept hierarchy, in one embodiment of the invention, set long, medium and short three property values.It is short that duration is less than or equal to 500 milliseconds of corresponding property values, is greater than 500 milliseconds and is less than in 1 second corresponding property value, being more than or equal to 1 second corresponding property value long.
(5) OCCURRENCE (the previous day whether report send out): be used for representing whether this warning information is reported the previous day and sent out.In one embodiment of the invention, in order to effectively classify, setting is, no two attributes.
(6) DECISION (whether being noise data): be used for representing whether warning information is noise data.In one embodiment of the invention, set be, no two property values.
As shown in Figure 4, to the data processing table formed, adopt the Decision Tree Algorithm of band Bayes node to classify, form denoising decision tree.Wherein, adopt in the process of Decision Tree Algorithm structure denoising decision tree of band Bayes node, first from data processing table, a part of data of random extraction, as a training data set samples, are expressed as { x
1, x
2..., x
n, in one embodiment of the invention, the data of random extraction 2/3 are as a training data set samples.Wherein, each data have 5 attributes, i.e. MSG_TYPE, DURATION, INTERVAL, OCCURRENCE, RECORD, are expressed as set attriubute_set.Each attribute has the property value of its correspondence.In one embodiment of the invention, the property value attribute1={a of MSG_TYPE
1, a
2, a
3, a
4}={ guard signal, switching signal, anticipating signal, alarm signal }; The property value attribute2={a of DURATION
1, a
2, a
3}={ is long, in, short }; The property value attribute3={a of INTERVAL
1, a
2, a
3}={ is long, in, short }; The property value attribute4={a of OCCURRENCE
1, a
2}={ is, no }; The property value attribute5={a of RECORD
1, a
2}={ has, nothing }, adopt the Decision Tree Algorithm of band Bayes node to classify on this basis, form denoising decision tree, the classification of generation is normal data, noise data two class, and as shown in Figure 5, concrete treatment step is as follows:
Step 11, extracts the property value of each attribute, utilizes the method for information gain-ratio, the information gain-ratio of each attribute in set of computations attribute_set from set attribute_set;
Step 12, selects information gain-ratio value to be maximum attribute root_attribute, is labeled as the intermediate node N of tree, and from set attribute_set delete property root_attribute;
Step 13, extracts each known property value a in attribute root_attribute
iif, property value a
iclassification unambiguity, so Bayes's node value is 0, and to grow a condition by node N be attribute root_attribute=a
ibranch, setting S
ifor attribute root_attribute=a in training data set samples
ia division; If S
ibe not empty, so on existing decision tree, add a leaf, be labeled as a classification in training data set samples; If property value a
iclassification has ambiguity, then turn to step 14;
Step 14, by certain principle to x
iclassify, if x
idetermine corresponding a certain classification c
j, then this type of is divided into; Otherwise, if x
ican not determine and assign to certain classification, but relevant to some classification, then according to prior imformation P (c
j) first it is placed in a certain class, then calculate P (x
i| c
j) and P (x
i), and calculate posterior probability according to these two values.If according to x
im (m<=5) individual attribute classify, and separate between attribute, then by x
ibe divided into x
i1, x
i2..., x
im, so P (x
i| c
j) can be expressed as: P (x
i1| c
j) × P (x
i2| c
j) × ... × P (x
im| c
j), thus can posterior probability be obtained:
according to the posterior probability obtained, by x
ibe divided in the class belonging to maximum posterior probability;
Step 15, chooses the f value of Bayes's node, and f=P (c
j| x
i);
Step 16, calculates successively, until do not have data message in set attriubute_set, final denoising decision tree generates.
As shown in Figure 6, be final denoising decision-tree model, wherein, n
i(i=0,1,2,3 ...) be attribute node, a
i(i=0,1,2,3 ...) be condition node, c
i(i=0,1,2,3 ...) be category node, 0/f is Bayes's node.Wherein, 0 value represents that this node does not carry out any calculating, directly according to condition a
ithe expression of next attribute node f value is turned to need computing function f.
After denoising decision tree generates, it is assessed, remaining 1/3 data are used to infer as test data, judge that it belongs to normal data or noise data, and itself and the classification belonging to original are compared, calculate the accuracy rate that it detects, improve the accuracy of denoising decision tree, and then improve the accuracy of denoising classifying rules.
For the ease of understanding, represent the denoising classifying rules of the warning information of being derived by denoising decision tree in table form.As shown in table 4, illustrate the denoising classifying rules of the warning information that part denoising decision tree is derived.
| MSG_TYPE | RECORD | OCCURRENCE | DURATION | INTERVAL | DECISION |
| Guard signal | Nothing | Be | Short | * | Noise data |
| * | Have | * | * | * | Noise data |
| …… | …… | …… | …… | …… | …… |
The denoising table of classification rules of table 4 warning information
Owing to combining with the remote measurement value in record of examination, SCADA in OMS, and summarize new decision attribute: OCCURRENCE, improve speed and the accuracy rate of denoising; Employ the decision Tree algorithms of band Bayes node, the defect that NB Algorithm cannot extract rule can be solved, the accuracy rate that decision Tree algorithms improves classification can be improved again.
In addition, warning information after denoising, filtering alarm noise information, iff directly showing these warning information, reasoning from logic is not carried out to it, cannot correctly reflect electrical network produced problem, therefore need to use knowledge rule to carry out reasoning to warning information.
Internal logical relationship is there is between the warning information that intelligent grid supporting system technology is uploaded, these logical relations are utilized to carry out reasoning to the warning information be associated, can the comprehensive warning information of reaching a conclusion property, clear tells management and running personnel electrical network produced problem, is convenient to management and running personnel and deals with in time, accurately.So need to carry out analyzing and processing to the warning information obtained, obtain electrical network alarm inference rule.Below this process is described in detail.
In warning information table after denoising, there are the multiple warning information much caused by same event.These warning information have consistance in time, and namely each information is sent out with time report or in succession reported in specific time interval and sends out.In one embodiment of the invention; according to the feature of electrical network; warning information is divided into protection, switch changed position, voltage out-of-limit, out-of-limit, meritorious out-of-limit, the switch trip of electric current; disconnecting link opening and closing; alarm 8 kinds, to send out with Times for the search of each warning information or (determine) report in succession at the appointed time according to on-the-spot operation rule, and has the warning information that identical plant stand (GROUPN) marks; form an alarm combination, and duplicate removal process is carried out to it.After duplicate removal, each alarm combination is deposited in warning information combination table, as shown in table 5, illustrate the partial information of warning information combination table.
| ID | Alarm is combined |
| 1 | { protection act, switch trip, alarm } |
| 2 | { switch changed position divides, voltage out-of-limit, voltage out-of-limit involution } |
| 3 | Alarm, and switch changed position, disconnecting link opening and closing, meritorious out-of-limit |
| …… | …… |
Table 5 warning information combination table
According to the warning information in above-mentioned warning information combination table and electric system knowledge, the alarm type belonging to it is judged for each alarm combination.Judge alarm type according to alarm combination, can find more alarm type, be not only common several alarm types, is convenient to extract more comprehensively electrical network warning information inference rule.
In one embodiment of the invention, warning information arranges different attributes respectively, has report to send out and do not report two attribute as protection, specifically as shown in table 6.
| 1. protect=report | 9. voltage out-of-limit=report is sent out |
| 2. protect=do not report and send out | 10. voltage out-of-limit=do not report and send out |
| 3. switch trip=report is sent out | 11. electric currents are out-of-limit=and report sends out |
| 4. switch trip=do not report and send out | 12. electric currents are out-of-limit=and do not report and send out |
| 5. switch changed position=report is sent out | 13. meritorious out-of-limit=report sends out |
| 6. switch changed position=do not report and send out | 14. meritorious out-of-limit=do not report and send out |
| 7. disconnecting link opening and closing=report is sent out | 15. alarm=reports are sent out |
| 8. disconnecting link opening and closing=do not report and send out | 16. alarms=do not report and send out |
The attribute list that table 6 warning information is corresponding
Alarm combination is carried out corresponding change, as { alarm, switch changed position by the attribute corresponding according to warning information in above-mentioned attribute list; disconnecting link opening and closing, meritorious out-of-limit become and { protect=do not report and send out, switch trip=do not report and send out; switch changed position=report is sent out; disconnecting link opening and closing=report is sent out, and voltage out-of-limit=do not report is sent out, and electric current is out-of-limit=and do not report and send out; out-of-limit=report of gaining merit is sent out; alarm=report is sent out }, then alarm new after changing form combination is combined with corresponding alarm type, form alarm affairs.Each alarm transaction packet is containing an alarm combination, and corresponding alarm type is combined in this alarm.Multiple alarm affairs form warning information transaction table, as shown in table 7, illustrate the partial information of warning information transaction table.
| TID | Alarm transaction item ({ alarm is combined, corresponding alarm type }) |
Table 7 warning information transaction table
Further analyzing and processing is carried out to warning information transaction table, electrical network alarm inference rule can be excavated.FP growth algorithm is most widely used in current Mining Frequent Itemsets Based algorithm, and does not need a kind of association rules mining algorithm of Candidate Set.But this algorithm also has some shortcomings.Its major defect has:
1, achievement and mining process all need to take a large amount of internal memories.When database is very large, or when the number of frequent 1-item collection in database is very large, travelling speed will greatly reduce; What is more, and may set due to the FP that cannot construct based on internal memory, this algorithm can not work effectively.
When 2, excavating large database, arithmetic speed is slow.
Not enough in order to overcome these, in one embodiment of the invention, use and improve FP growth algorithm, do not need on the basis of the advantage producing marquis's set of choices at succession FP growth algorithm, database is carried out the item always scanning for several times of frequent 1-item collection, each scanning obtains the database subset of the item of each frequent 1-item collection respectively.Then use FP growth algorithm to carry out constrained frequent item sets mining to every database subset respectively, obtain the frequent item set of the item containing each frequent 1-item collection, finally these frequent item sets are combined all frequent item sets just obtaining whole database.Concrete treatment step is as follows:
Step 21, the warning information transaction table D in scan database, finds out the set of candidate 1-item collection, and obtains their support counting (frequency).Then, successively decrease according to support counting and arrange the every of candidate 1-item collection, obtain the set F of the candidate 1-item collection that support counting successively decreases.Support in F is less than the entry deletion of minimum support threshold values min_sup, obtains the set L of frequent 1-item collection, L={I
1, I
2..., I
m, wherein, I
1support the highest, I
msupport minimum.In one embodiment of the invention, minimum support threshold values min_sup is set to 1%, does not miss special alarm affairs as far as possible, improves the accuracy of alarm and the comprehensive of electrical network alarm inference rule.
Step 22, warning information transaction table D again in scan database, item support being less than minimum support threshold values min_sup is deleted from each affairs, is rearranged by the item in each affairs degressively according to every support counting, obtains the warning information transaction table D' after processing.
Step 23, according to the every support counting in frequent 1-item collection L, ascendingly constructs every database subset successively, and utilizes FP growth algorithm to carry out constrained frequent item sets mining to it.Wherein, ascendingly construct every database subset successively and comprise for each I in L
i(i=m, m-1 ..., 1) and carry out the process of following steps:
Step 231, the warning information transaction table D' in scan database after process, therefrom extracts all containing item I
iaffairs, delete support in these affairs and be less than the item of this support, obtain an I
idatabase subset D
i.
Step 232, to database subset D
i, utilize FP growth algorithm to carry out comprising an I
iconstraint frequent item set mining, its mining process comprises the steps:
Step 2321, utilizes database subset D
i, structure FP tree, and create item head table HT.Wherein, FP tree is by each Transaction Information item in Transaction Information table according to after support sequence, and it is in the tree of root node with sky that the data item in each affairs is inserted into one successively by descending, records the support that this node occurs at each node place simultaneously.In one embodiment of the invention, during structure FP tree, this database subset D
iin the item of each affairs according to the order process in frequent 1-item collection L.Therefore, that last in item head table HT indicates is exactly item I
isupport counting and node chain information.
Step 2322, by last information indicated in item head table HT, constructs this conditional pattern base, constructs its condition FP and sets, and this condition FP tree is excavated the constraint frequent item collection CL comprising this
i, complete at database subset D
ion constraint frequent item set mining.Wherein, conditional pattern base is the set comprising the prefix path occurred together with suffix pattern in FP tree.
Step 24, as the constraint frequent item collection CL of items all in L
iafter excavating out successively, merge these constraint frequent item collection, namely get these constraint frequent item collection CL
iunion, obtain all frequent item sets of the warning information transaction table D in database, terminate mining process.
After FP growth algorithm, produce the frequent item set in alarm transaction table, namely meet all alarm transaction item of minimum support threshold values min_sup.Association rule mining is carried out to the alarm transaction item (frequent item set) meeting minimum support threshold values min_sup, just can obtain electrical network alarm inference rule.
When finding out frequent item set from the warning information transaction table in database, producing Strong association rule by it, needing Strong association rule to meet minimum support and min confidence.Wherein, for degree of confidence, can obtain with following formula:
Wherein, conditional probability item collection support counting represents, sup_cnt (A ∪ B) is the number of transactions comprising item collection A ∪ B, and sup_cnt (A) is the number of transactions comprising item collection A, and according to formula, correlation rule can produce as follows:
(1) for each frequent item set l, all nonvoid subsets of l are produced.
(2) for each nonvoid subset s of l, if
then export rule
wherein, min_conf is min confidence threshold values.
In one embodiment of the invention, A is the alarm combination in frequent item set, and B is the alarm classification in frequent item set, and min_conf is set to 100%, can be derived by above-mentioned correlation rule
as
namely electrical network alarm inference rule is obtained.
In sum, intelligent alarm analytical approach towards dispatching of power netwoks provided by the present invention, by by the record of examination in warning information and OMS, and the remote measurement value in SCADA combines, conclude the decision attribute made new advances, reduce the denoising time, and improve denoising accuracy rate, use the decision Tree algorithms of band Bayes node to carry out denoising to warning information, the defect that NB Algorithm cannot extract rule can be solved, the accuracy rate that decision Tree algorithms improves classification can be improved again.In addition, FP growth algorithm generation frequent item set is improved by using, recycling correlation rule produces electrical network alarm inference rule, improve the comprehensive of efficiency and alarm type, improve the rejection rate to noise data in warning information, excavate the incidence relation between warning information, obtain comprehensive electrical network alarm inference rule.
Above the intelligent alarm analytical approach towards dispatching of power netwoks provided by the present invention is described in detail.To those skilled in the art, to any apparent change that it does under the prerequisite not deviating from connotation of the present invention, all by formation to infringement of patent right of the present invention, corresponding legal liabilities will be born.
Claims (9)
1., towards an intelligent alarm analytical approach for dispatching of power netwoks, it is characterized in that comprising the steps:
Extract the warning information in electrical network, pre-service is carried out to described warning information;
Carry out denoising to through pretreated warning information, and obtain denoising classifying rules;
Warning information after denoising is concluded, duplicate removal, generate alarm combination, for the alarm type that each alarm combination judges belonging to it;
The alarm combination alarm type corresponding with it is carried out to the excavation of correlation rule, obtain electrical network alarm inference rule.
2. intelligent alarm analytical approach as claimed in claim 1, is characterized in that:
When carrying out pre-service to warning information, first warning information extracted, change, load process, remove incoherent warning information, then warning information is associated with record of examination, threshold value according to setting removes noise data, associates, warning information according to the change information of remote measurement value with the remote measurement value in SCADA, judge electrical network warning information whether mistake, remove the warning information of mistake.
3. intelligent alarm analytical approach as claimed in claim 1, is characterized in that:
When carrying out denoising to warning information, analyze through pretreated warning information, process, being formed to meet adopts the decision Tree algorithms of band Bayes node to carry out the data processing table of classifying, therefrom a part of data of random extraction are carried out classification and are formed denoising decision tree, judge whether warning information is noise data, and denoising decision tree is changed into denoising classifying rules.
4. intelligent alarm analytical approach as claimed in claim 3, is characterized in that:
After denoising decision tree generates, use remaining data to infer as test data, judge that it belongs to normal data or noise data, and itself and the classification belonging to original are compared, the accuracy rate that assessment detects.
5. intelligent alarm analytical approach as claimed in claim 1, is characterized in that:
When carrying out the excavation of correlation rule to the alarm combination Alarm Classification corresponding with it, first attribute corresponding with warning information for alarm combination being combined, generating new alarm combination; Again new alarm combination is combined with corresponding alarm type, form alarm affairs; The warning information transaction table of multiple alarm affairs formation is carried out to the excavation of frequent item set, association rule mining is carried out to the frequent item set meeting minimum support threshold values, obtains electrical network alarm inference rule.
6. intelligent alarm analytical approach as claimed in claim 5, when it is characterized in that the warning information transaction table formed multiple alarm affairs carries out the excavation of frequent item set, comprises the steps:
Step 21, the warning information transaction table in scan database, finds out the set of candidate, and obtains their support counting; Successively decrease according to support counting and arrange the every of candidate, obtain gathering F, support in set F is less than the entry deletion of minimum support threshold values, obtains the set L of frequent item set;
Step 22, warning information transaction table again in scan database, item support being less than minimum support threshold values is deleted from each affairs, and successively decreasing according to every support counting rearranges the item in each affairs, obtains the warning information transaction table after processing;
Step 23, according to every support counting in set L, ascendingly constructs every database subset successively, and utilizes FP growth algorithm to carry out constrained frequent item sets mining to it;
Step 24, after the constraint frequent item collection of items all in set L excavates out successively, merges all constraint frequent item collection, obtains all frequent item sets of warning information transaction table.
7. intelligent alarm analytical approach as claimed in claim 6, is characterized in that the ascending step constructing every database subset successively comprises further:
Step 231, the warning information transaction table in scan database after process, therefrom extracts all containing item I
iaffairs, delete support and be less than the item of this support, obtain an I
idatabase subset;
Step 232, to database subset, utilizes FP growth algorithm to carry out comprising an I
iconstraint frequent item set mining.
8. intelligent alarm analytical approach as claimed in claim 7, is characterized in that utilizing FP growth algorithm to carry out comprising an I
iconstraint frequent item set mining comprise the steps:
Step 2321, utilizes database subset, and structure FP tree, and creates item head table, and that last of wherein said item head table indicates is an I
isupport counting and node chain information;
Step 2322, by last information indicated of described item head table, construct this conditional pattern base, construct its condition FP and set, this condition FP tree is excavated the constraint frequent item collection comprising this, completes the constraint frequent item set mining on database subset.
9. intelligent alarm analytical approach as claimed in claim 5, is characterized in that the frequent item set to meeting minimum support threshold values carries out association rule mining, obtaining electrical network alarm inference rule and comprising the steps:
Computing formula according to degree of confidence:
Obtain correlation rule:
(1) for each frequent item set l, all nonvoid subsets of l are produced;
(2) for each nonvoid subset s of l, if
then export rule
Alarm combination in frequent item set is set to A, and alarm classification is set to B, and min_conf is set to 100%, derives electrical network alarm inference rule be by described correlation rule:
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410433102.2A CN104239437B (en) | 2014-08-28 | 2014-08-28 | A kind of intelligent alarm analysis method towards dispatching of power netwoks |
| PCT/CN2014/091711 WO2016029570A1 (en) | 2014-08-28 | 2014-11-19 | Intelligent alert analysis method for power grid scheduling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410433102.2A CN104239437B (en) | 2014-08-28 | 2014-08-28 | A kind of intelligent alarm analysis method towards dispatching of power netwoks |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104239437A true CN104239437A (en) | 2014-12-24 |
| CN104239437B CN104239437B (en) | 2017-12-15 |
Family
ID=52227497
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410433102.2A Expired - Fee Related CN104239437B (en) | 2014-08-28 | 2014-08-28 | A kind of intelligent alarm analysis method towards dispatching of power netwoks |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN104239437B (en) |
| WO (1) | WO2016029570A1 (en) |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105427043A (en) * | 2015-11-20 | 2016-03-23 | 江苏省电力公司扬州供电公司 | Improved nearest neighbor algorithm-based power grid alarm analysis method |
| CN106022950A (en) * | 2016-05-06 | 2016-10-12 | 中国电力科学研究院 | Power distribution network secondary equipment type identification method and system |
| CN106228244A (en) * | 2016-07-12 | 2016-12-14 | 深圳大学 | A kind of energy based on self adaptation association rule mining depolymerizes method |
| CN106452825A (en) * | 2016-07-20 | 2017-02-22 | 国网江苏省电力公司南京供电公司 | Power distribution and utilization communication network alarm correlation analysis method based on improved decision tree |
| CN106878038A (en) * | 2015-12-10 | 2017-06-20 | 华为技术有限公司 | Fault Locating Method and device in a kind of communication network |
| CN107247995A (en) * | 2016-09-29 | 2017-10-13 | 上海交通大学 | Transmission line of electricity running status association rule mining and Forecasting Methodology based on Bayesian model |
| CN107247970A (en) * | 2017-06-23 | 2017-10-13 | 国家质量监督检验检疫总局信息中心 | A kind of method for digging and device of commodity qualification rate correlation rule |
| CN107291716A (en) * | 2016-03-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | A kind of link data method of calibration and device |
| CN107391515A (en) * | 2016-05-17 | 2017-11-24 | 李明轩 | Power system index analysis method based on Association Rule Analysis |
| CN107967199A (en) * | 2017-12-05 | 2018-04-27 | 广东电网有限责任公司东莞供电局 | A kind of power equipment temperature pre-warning analysis method based on association rule mining |
| CN109191023A (en) * | 2018-11-07 | 2019-01-11 | 广东电网有限责任公司 | A kind of power grid warning information immediate processing method and device |
| CN109522388A (en) * | 2018-11-02 | 2019-03-26 | 中国联合网络通信集团有限公司 | A kind of creation method and device of intelligence worksheet processing rule |
| CN109739846A (en) * | 2018-12-27 | 2019-05-10 | 国电南瑞科技股份有限公司 | A kind of electric network data mass analysis method |
| CN109753526A (en) * | 2018-12-28 | 2019-05-14 | 四川新网银行股份有限公司 | A kind of device and method that warning information analysis is inquired based on timing similarity |
| CN111062503A (en) * | 2019-12-19 | 2020-04-24 | 国网山东省电力公司泰安供电公司 | Power grid monitoring alarm processing method, system, terminal and storage medium |
| CN111143428A (en) * | 2019-11-30 | 2020-05-12 | 贵州电网有限责任公司 | Protection abnormity alarm processing method based on correlation analysis method |
| CN111832769A (en) * | 2019-09-24 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method and device for ordering vehicle-entering points and information |
| CN111898674A (en) * | 2020-07-29 | 2020-11-06 | 中国电力科学研究院有限公司 | Network fault positioning model training and identifying method, device, equipment and medium |
| CN112286987A (en) * | 2020-10-21 | 2021-01-29 | 国网电力科学研究院武汉南瑞有限责任公司 | Electric power internet of things abnormal alarm compression method based on Apriori algorithm |
| CN112699005A (en) * | 2020-12-30 | 2021-04-23 | 网宿科技股份有限公司 | Server hardware fault monitoring method, electronic equipment and storage medium |
| CN113064934A (en) * | 2021-03-26 | 2021-07-02 | 安徽继远软件有限公司 | Fault association rule mining method and system for sensing layer of power sensor network |
| CN113282686A (en) * | 2021-06-03 | 2021-08-20 | 光大科技有限公司 | Method and device for determining association rule of unbalanced sample |
| CN113448763A (en) * | 2021-07-16 | 2021-09-28 | 广东电网有限责任公司 | Dynamic expansion grouping alarm service method for full life cycle management |
| CN113591393A (en) * | 2021-08-10 | 2021-11-02 | 国网河北省电力有限公司电力科学研究院 | Fault diagnosis method, device, equipment and storage medium of intelligent substation |
| CN113836196A (en) * | 2021-09-08 | 2021-12-24 | 国网江苏省电力有限公司 | Power grid undefined event type identification method and system |
| CN114024829A (en) * | 2021-10-26 | 2022-02-08 | 广东电网有限责任公司 | Fault repairing method, device, equipment and storage medium of power communication network |
| CN114338746A (en) * | 2021-12-30 | 2022-04-12 | 以萨技术股份有限公司 | Analysis early warning method and system for data collection of Internet of things equipment |
| CN114358062A (en) * | 2021-12-23 | 2022-04-15 | 河南大学 | Yellow river bank dam dangerous case identification method based on formal concept analysis |
| WO2022111659A1 (en) * | 2020-11-30 | 2022-06-02 | 中兴通讯股份有限公司 | Warning method, apparatus and device, and storage medium |
| CN116246445A (en) * | 2023-03-14 | 2023-06-09 | 国网湖北省电力有限公司孝感供电公司 | Knowledge-graph-based warehouse safety multi-source Internet-of-things data early warning method |
Families Citing this family (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600063B (en) * | 2016-12-16 | 2022-03-18 | 东北电力大学 | Space load prediction method based on fuzzy rough set theory and spatio-temporal information |
| CN107835087B (en) * | 2017-09-14 | 2022-09-02 | 北京科东电力控制系统有限责任公司 | Automatic extraction method of alarm rule of safety equipment based on frequent pattern mining |
| CN108880915B (en) * | 2018-08-20 | 2023-03-24 | 全球能源互联网研究院有限公司 | Electric power information network safety alarm information false alarm determination method and system |
| CN109256114B (en) * | 2018-08-30 | 2022-12-13 | 国网浙江省电力有限公司杭州供电公司 | Voice alarm method for power grid monitoring and dispatching |
| CN111241145A (en) * | 2018-11-28 | 2020-06-05 | 中国移动通信集团浙江有限公司 | Self-healing rule mining method and device based on big data |
| CN109597836B (en) * | 2018-11-29 | 2023-06-27 | 武汉大学 | Communication equipment alarm association rule mining method based on weighting matrix |
| CN109582714B (en) * | 2018-12-03 | 2023-07-25 | 中电万维信息技术有限责任公司 | Government affair item data processing method based on time attenuation association |
| CN109783310A (en) * | 2018-12-14 | 2019-05-21 | 深圳壹账通智能科技有限公司 | The Dynamic and Multi dimensional method for safety monitoring and its monitoring device of information technoloy equipment |
| CN109767108B (en) * | 2018-12-31 | 2023-03-24 | 国网安徽省电力有限公司 | Accurate analysis method for line state of power dispatching system |
| CN109871305B (en) * | 2019-01-18 | 2022-11-04 | 深圳壹账通智能科技有限公司 | Alarm information processing method and device, computer equipment and storage medium |
| CN109902111B (en) * | 2019-03-20 | 2023-04-07 | 水利部交通运输部国家能源局南京水利科学研究院 | Earth and rockfill dam safety information mining method |
| CN109993443B (en) * | 2019-04-04 | 2023-01-31 | 北京科东电力控制系统有限责任公司 | Method for mining power grid equipment operation state space-time situation perception model |
| CN112054913B (en) * | 2019-06-05 | 2023-07-18 | 厦门网宿有限公司 | Data monitoring system and method |
| CN110399262B (en) * | 2019-06-17 | 2022-09-27 | 平安科技（深圳）有限公司 | Operation and maintenance monitoring alarm convergence method and device, computer equipment and storage medium |
| CN110932881B (en) * | 2019-10-16 | 2022-07-26 | 国网浙江省电力有限公司温州供电公司 | Equipment portrait method based on monitoring alarm signal |
| CN111177223A (en) * | 2019-12-27 | 2020-05-19 | 国网北京市电力公司 | Voltage abnormity analysis method and device and electronic equipment |
| CN111125268B (en) * | 2019-12-27 | 2024-01-30 | 南京亚信软件有限公司 | Network alarm analysis model creation method, alarm analysis method and device |
| CN111258798A (en) * | 2020-01-14 | 2020-06-09 | 平安壹钱包电子商务有限公司 | Fault positioning method and device for monitoring data, computer equipment and storage medium |
| CN111626792A (en) * | 2020-06-01 | 2020-09-04 | 长沙理工大学 | Technology for accurately portraying load storage of comprehensive energy source in distribution network |
| CN112131265A (en) * | 2020-07-10 | 2020-12-25 | 国网河北省电力有限公司雄安新区供电公司 | Contactor electric service life bad data detection method based on data mining |
| CN112148772A (en) * | 2020-09-24 | 2020-12-29 | 创新奇智(成都)科技有限公司 | Alarm root cause identification method, device, equipment and storage medium |
| CN112183656B (en) * | 2020-10-12 | 2023-09-29 | 国网新疆电力有限公司 | SCADA data frequent item set mining method in power grid fault |
| CN112347225B (en) * | 2020-10-19 | 2022-09-16 | 云南电网有限责任公司 | Processing method for transmitting data information across isolated network gate |
| CN113312800B (en) * | 2021-06-28 | 2023-01-17 | 西安热工研究院有限公司 | Data simulation method for power station industrial control safety target range |
| CN113536312A (en) * | 2021-07-28 | 2021-10-22 | 工银科技有限公司 | Alarm information processing method and device |
| CN113759165A (en) * | 2021-08-13 | 2021-12-07 | 贵州电网有限责任公司遵义供电局 | Power distribution network fault power failure perception analysis method and equipment |
| CN113743512A (en) * | 2021-09-07 | 2021-12-03 | 上海观安信息技术股份有限公司 | Autonomous learning judgment method and system for safety alarm event |
| CN114185957B (en) * | 2021-11-16 | 2024-04-05 | 国网辽宁省电力有限公司 | Intelligent mining method suitable for power big data service requirements |
| CN114257414A (en) * | 2021-11-25 | 2022-03-29 | 国网山东省电力公司日照供电公司 | Intelligent network security duty method and system |
| CN114500011B (en) * | 2022-01-13 | 2023-12-05 | 中国电子科技网络信息安全有限公司 | Auxiliary decision-making method based on behavior baseline anomaly analysis and event arrangement |
| CN114553560B (en) * | 2022-02-25 | 2024-01-30 | 国网山东省电力公司临沂供电公司 | Intelligent network security alarm analysis platform based on big data technology |
| CN115882938B (en) * | 2023-02-20 | 2023-05-16 | 中通服建设有限公司 | Optical network fault monitoring system |
| CN116054416B (en) * | 2023-03-15 | 2023-09-22 | 扬州康德电气有限公司 | Intelligent monitoring operation and maintenance management system based on Internet of things |
| CN116629709B (en) * | 2023-07-21 | 2023-10-20 | 国网山东省电力公司青岛市即墨区供电公司 | Intelligent analysis alarm system of power supply index |
| CN117094006B (en) * | 2023-10-20 | 2024-02-23 | 湖南三湘银行股份有限公司 | Security event root cause analysis method and system based on artificial intelligence algorithm |
| CN117806916A (en) * | 2024-02-29 | 2024-04-02 | 中国人民解放军国防科技大学 | Multi-unit server lightweight alarm correlation mining and converging method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208049A (en) * | 2013-04-25 | 2013-07-17 | 国家电网公司 | Quick accident analysis method and system for abnormal alarm |
| CN103489136A (en) * | 2013-09-30 | 2014-01-01 | 陕西省地方电力（集团）有限公司 | SCADA warning message processing method and device based on expert system tool |
| CN103514516A (en) * | 2013-09-27 | 2014-01-15 | 国家电网公司 | Method for obtaining power grid device failure information based on calculation of multi-source redundant information |
| CN103700031A (en) * | 2013-12-19 | 2014-04-02 | 国家电网公司 | Electric power warning information publishing method in regulation and control integration mode |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101937447B (en) * | 2010-06-07 | 2012-05-23 | 华为技术有限公司 | Alarm association rule mining method, and rule mining engine and system |
| CN102098175B (en) * | 2011-01-26 | 2015-07-01 | 浪潮通信信息系统有限公司 | Alarm association rule obtaining method of mobile internet |
-
2014
- 2014-08-28 CN CN201410433102.2A patent/CN104239437B/en not_active Expired - Fee Related
- 2014-11-19 WO PCT/CN2014/091711 patent/WO2016029570A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208049A (en) * | 2013-04-25 | 2013-07-17 | 国家电网公司 | Quick accident analysis method and system for abnormal alarm |
| CN103514516A (en) * | 2013-09-27 | 2014-01-15 | 国家电网公司 | Method for obtaining power grid device failure information based on calculation of multi-source redundant information |
| CN103489136A (en) * | 2013-09-30 | 2014-01-01 | 陕西省地方电力（集团）有限公司 | SCADA warning message processing method and device based on expert system tool |
| CN103700031A (en) * | 2013-12-19 | 2014-04-02 | 国家电网公司 | Electric power warning information publishing method in regulation and control integration mode |
Non-Patent Citations (3)
| Title |
|---|
| 吴简: "面向业务的基于模糊关联规则挖掘的网络故障诊断系统", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 张兴: "地区电网二次报警信息智能处理系统的研究与开发", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 晁进: "基于数据挖掘技术的电网智能报警系统的研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105427043A (en) * | 2015-11-20 | 2016-03-23 | 江苏省电力公司扬州供电公司 | Improved nearest neighbor algorithm-based power grid alarm analysis method |
| CN106878038A (en) * | 2015-12-10 | 2017-06-20 | 华为技术有限公司 | Fault Locating Method and device in a kind of communication network |
| CN107291716A (en) * | 2016-03-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | A kind of link data method of calibration and device |
| CN107291716B (en) * | 2016-03-30 | 2020-07-21 | 阿里巴巴集团控股有限公司 | Link data checking method and device |
| CN106022950A (en) * | 2016-05-06 | 2016-10-12 | 中国电力科学研究院 | Power distribution network secondary equipment type identification method and system |
| CN106022950B (en) * | 2016-05-06 | 2020-12-18 | 中国电力科学研究院有限公司 | Power distribution network secondary equipment type identification method and system |
| CN107391515A (en) * | 2016-05-17 | 2017-11-24 | 李明轩 | Power system index analysis method based on Association Rule Analysis |
| CN106228244A (en) * | 2016-07-12 | 2016-12-14 | 深圳大学 | A kind of energy based on self adaptation association rule mining depolymerizes method |
| CN106452825A (en) * | 2016-07-20 | 2017-02-22 | 国网江苏省电力公司南京供电公司 | Power distribution and utilization communication network alarm correlation analysis method based on improved decision tree |
| CN107247995A (en) * | 2016-09-29 | 2017-10-13 | 上海交通大学 | Transmission line of electricity running status association rule mining and Forecasting Methodology based on Bayesian model |
| CN107247970A (en) * | 2017-06-23 | 2017-10-13 | 国家质量监督检验检疫总局信息中心 | A kind of method for digging and device of commodity qualification rate correlation rule |
| CN107967199A (en) * | 2017-12-05 | 2018-04-27 | 广东电网有限责任公司东莞供电局 | A kind of power equipment temperature pre-warning analysis method based on association rule mining |
| CN107967199B (en) * | 2017-12-05 | 2019-11-08 | 广东电网有限责任公司东莞供电局 | A kind of power equipment temperature pre-warning analysis method based on association rule mining |
| CN109522388A (en) * | 2018-11-02 | 2019-03-26 | 中国联合网络通信集团有限公司 | A kind of creation method and device of intelligence worksheet processing rule |
| CN109191023A (en) * | 2018-11-07 | 2019-01-11 | 广东电网有限责任公司 | A kind of power grid warning information immediate processing method and device |
| CN109739846A (en) * | 2018-12-27 | 2019-05-10 | 国电南瑞科技股份有限公司 | A kind of electric network data mass analysis method |
| CN109753526A (en) * | 2018-12-28 | 2019-05-14 | 四川新网银行股份有限公司 | A kind of device and method that warning information analysis is inquired based on timing similarity |
| CN111832769A (en) * | 2019-09-24 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method and device for ordering vehicle-entering points and information |
| CN111143428A (en) * | 2019-11-30 | 2020-05-12 | 贵州电网有限责任公司 | Protection abnormity alarm processing method based on correlation analysis method |
| CN111062503A (en) * | 2019-12-19 | 2020-04-24 | 国网山东省电力公司泰安供电公司 | Power grid monitoring alarm processing method, system, terminal and storage medium |
| CN111062503B (en) * | 2019-12-19 | 2023-08-18 | 国网山东省电力公司泰安供电公司 | Power grid monitoring alarm processing method, system, terminal and storage medium |
| CN111898674A (en) * | 2020-07-29 | 2020-11-06 | 中国电力科学研究院有限公司 | Network fault positioning model training and identifying method, device, equipment and medium |
| CN112286987B (en) * | 2020-10-21 | 2022-04-29 | 国网电力科学研究院武汉南瑞有限责任公司 | Electric power internet of things abnormal alarm compression method based on Apriori algorithm |
| CN112286987A (en) * | 2020-10-21 | 2021-01-29 | 国网电力科学研究院武汉南瑞有限责任公司 | Electric power internet of things abnormal alarm compression method based on Apriori algorithm |
| WO2022111659A1 (en) * | 2020-11-30 | 2022-06-02 | 中兴通讯股份有限公司 | Warning method, apparatus and device, and storage medium |
| CN112699005A (en) * | 2020-12-30 | 2021-04-23 | 网宿科技股份有限公司 | Server hardware fault monitoring method, electronic equipment and storage medium |
| CN113064934A (en) * | 2021-03-26 | 2021-07-02 | 安徽继远软件有限公司 | Fault association rule mining method and system for sensing layer of power sensor network |
| CN113064934B (en) * | 2021-03-26 | 2023-12-08 | 安徽继远软件有限公司 | Power sensing network perception layer fault association rule mining method and system |
| CN113282686A (en) * | 2021-06-03 | 2021-08-20 | 光大科技有限公司 | Method and device for determining association rule of unbalanced sample |
| CN113282686B (en) * | 2021-06-03 | 2023-11-07 | 光大科技有限公司 | Association rule determining method and device for unbalanced sample |
| CN113448763B (en) * | 2021-07-16 | 2022-07-26 | 广东电网有限责任公司 | Dynamic expansion grouping alarm service method for full life cycle management |
| CN113448763A (en) * | 2021-07-16 | 2021-09-28 | 广东电网有限责任公司 | Dynamic expansion grouping alarm service method for full life cycle management |
| CN113591393A (en) * | 2021-08-10 | 2021-11-02 | 国网河北省电力有限公司电力科学研究院 | Fault diagnosis method, device, equipment and storage medium of intelligent substation |
| CN113591393B (en) * | 2021-08-10 | 2024-05-31 | 国网河北省电力有限公司电力科学研究院 | Fault diagnosis method, device, equipment and storage medium of intelligent substation |
| CN113836196A (en) * | 2021-09-08 | 2021-12-24 | 国网江苏省电力有限公司 | Power grid undefined event type identification method and system |
| CN114024829A (en) * | 2021-10-26 | 2022-02-08 | 广东电网有限责任公司 | Fault repairing method, device, equipment and storage medium of power communication network |
| CN114358062A (en) * | 2021-12-23 | 2022-04-15 | 河南大学 | Yellow river bank dam dangerous case identification method based on formal concept analysis |
| CN114338746A (en) * | 2021-12-30 | 2022-04-12 | 以萨技术股份有限公司 | Analysis early warning method and system for data collection of Internet of things equipment |
| CN116246445A (en) * | 2023-03-14 | 2023-06-09 | 国网湖北省电力有限公司孝感供电公司 | Knowledge-graph-based warehouse safety multi-source Internet-of-things data early warning method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104239437B (en) | 2017-12-15 |
| WO2016029570A1 (en) | 2016-03-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104239437A (en) | Power-network-dispatching-oriented intelligent warning analysis method | |
| CN106019084B (en) | Medium-voltage power distribution network disconnection fault diagnosis method based on power distribution and utilization data correlation | |
| CN107609569B (en) | Power distribution network ground fault positioning method based on multi-dimensional feature vectors | |
| CN110958136A (en) | Deep learning-based log analysis early warning method | |
| CN103455563B (en) | A kind of data digging method being applicable to intelligent substation integrated monitoring system | |
| Zhang et al. | Time series anomaly detection for smart grids: A survey | |
| CN104363106B (en) | A kind of communicating for power information fault pre-alarming analysis method based on big data technology | |
| Zhao et al. | Hierarchical anomaly detection and multimodal classification in large-scale photovoltaic systems | |
| CN103902816A (en) | Electrification detection data processing method based on data mining technology | |
| CN110807550B (en) | Distribution transformer overload recognition and early warning method based on neural network and terminal equipment | |
| CN104616092A (en) | Distributed log analysis based distributed mode handling method | |
| CN104463709A (en) | Substation alarm information processing method based on decision trees | |
| CN107798395A (en) | A kind of power grid accident signal automatic diagnosis method and system | |
| CN109615004A (en) | A kind of anti-electricity-theft method for early warning of multisource data fusion | |
| CN115271263B (en) | Power equipment defect early warning method, system and medium based on improved association rule | |
| CN104731953A (en) | R-based building method of data preprocessing system | |
| Kezunovic | Translational knowledge: From collecting data to making decisions in a smart grid | |
| CN105930424A (en) | Method for realizing online asynchronous acquisition and intelligent mining of power distribution network data | |
| CN105528671A (en) | Power grid multidimensional sensing and safety assessment system and method on the basis of big data | |
| Gholami et al. | Comparative analysis of ml techniques for data-driven anomaly detection, classification and localization in distribution system | |
| Andrade et al. | Data-driven anomaly detection and event log profiling of SCADA alarms | |
| CN110334127A (en) | A kind of distribution network line fault law mining method, system and storage medium | |
| Bionda et al. | Covid-19 and electricity demand: focus on Milan and Brescia distribution grids | |
| CN104978837A (en) | Alarm system orienting user-side transformer station and realization method thereof | |
| Khediri et al. | Deep-belief network based prediction model for power outage in smart grid |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160331 Address after: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant after: State Grid Corporation of China Applicant after: Beijing Kedong Power Control System Co., Ltd. Applicant after: State Grid Heilongjiang Electric Power Co., Ltd. Applicant after: State Grid Zhejiang Electric Power Company Address before: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant before: State Grid Corporation of China Applicant before: Beijing Kedong Power Control System Co., Ltd. Applicant before: State Grid Heilongjiang Electric Power Co., Ltd. |
|
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Shang Xuewei Inventor after: Li Bing Inventor after: Cui Xu Inventor after: Ma Zhongjia Inventor after: Fu Lisu Inventor after: Guo Jiancheng Inventor after: Li Yetian Inventor after: Chen Cuan Inventor after: Wang Zan Inventor after: Tian Shigang Inventor after: Di Yong Inventor after: Li Shilun Inventor after: Zhang Liang Inventor before: Shang Xuewei Inventor before: Cui Xu Inventor before: Ma Zhongjia Inventor before: Fu Lisu Inventor before: Li Yetian Inventor before: Chen Cuan Inventor before: Wang Zan Inventor before: Tian Shigang Inventor before: Di Yong Inventor before: Li Shilun Inventor before: Zhang Liang Inventor before: Li Bing |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171215 Termination date: 20190828 |