CN109188330A - A kind of fault diagnosis method and system of distribution network failure indicator - Google Patents
A kind of fault diagnosis method and system of distribution network failure indicator Download PDFInfo
- Publication number
- CN109188330A CN109188330A CN201811001810.3A CN201811001810A CN109188330A CN 109188330 A CN109188330 A CN 109188330A CN 201811001810 A CN201811001810 A CN 201811001810A CN 109188330 A CN109188330 A CN 109188330A
- Authority
- CN
- China
- Prior art keywords
- fault detector
- ratio
- working
- module
- history
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application discloses a kind of fault diagnosis method and system of distribution network failure indicator, the described method includes: the corresponding relationship of the attribute value of setting fault detector and each module under each working, history working is acquired, configuration node sample set calculates information gain-ratio;The corresponding module of maximum information gain-ratio is selected as current layer root node, selects the main split of current layer root node;The corresponding module of maximum information gain-ratio is selected from remaining information ratio of profit increase and forms new current layer root node, selects the main split of new current layer root node;If there is no remaining information ratio of profit increase, decision tree is generated;Acquisition work at present operating status obtains the diagnostic result of fault detector according to decision tree in real time.The failure of fault detector can be found in time in fault detector itself fail using preceding method or system, the specific module that indicating fault breaks down can be determined according to the branch of decision tree, reduce maintenance engineering amount.
Description
Technical field
This application involves distribution network failure positioning device fields more particularly to a kind of failure of distribution network failure indicator to examine
Disconnected method and system.
Background technique
With electric utility fast development, the power circuit and power grid of power distribution network are more and more intensive, lead to the event of power distribution network
Barrier also can be more and more.The failure of power distribution network can bring huge economic loss, while can also give people's lives band
It is greatly inconvenient to come.Therefore, in order to ensure normal power supply, it is necessary to ensure the unimpeded of distribution.Pacified at present using fault detector
In power distribution network, timely to indicate the failure of power distribution network, so that maintenance personnel can timely have found failure and solve
Failure.
Fault detector in power distribution network monitors in real time and acquires the line status of power distribution network, and by the route shape of acquisition
The information of state is sent to the main website center of power distribution network, once the line failure of power distribution network, the information of line status can occur
Abnormal, main website center counts the line state information being abnormal, analyze and topology calculates, determine failure cause and therefore
Barrier region repairs in time so that staff be guided rapidly and accurately to find abort situation.Therefore, fault detector is to matching
Power grid is critically important, once fault detector itself fail, will be unable to the failure of monitoring power distribution network.But fault detector
It itself can not be diagnosed to be the failure of itself, in the prior art, do not send the route of power distribution network for a long time in fault detector usually
In the case where status information to main website center, determine that failure itself has occurred in this fault detector, and determining fault detector
After breaking down, the location information that the service personnel of fault detector pre-sets according to main website center is preceding that failure is gone to refer to
Position where showing device is gone to overhaul on the spot, by overhauling on the spot, determines the specific module that indicating fault breaks down.
But inventor has found in the research process of the application, and in the prior art, fault detector itself fail
When, when main website center does not receive the information of line status for a long time, it just can determine that failure has occurred in fault detector itself,
Lead to not the failure for finding fault detector in time, meanwhile, it just can determine that indicating fault broke down by overhauling on the spot
Specific module causes maintenance engineering amount larger.
Summary of the invention
The fault diagnosis method and system of a kind of distribution network failure indicator provided by the present application, to solve in the prior art
When fault detector itself fail, the failure of fault detector can not be found in time, and can not determine that indicating fault is sent out
The specific module of raw failure, the problem for causing maintenance engineering amount larger.
In a first aspect, the embodiment of the present application provides a kind of method for diagnosing faults of distribution network failure indicator, the failure
Diagnostic method includes:
1) be arranged modules in fault detector and fault detector respectively various workings under category
The corresponding relationship of property value;
2) the history working for acquiring modules in the fault detector and fault detector, according to described
The corresponding relationship of various workings and attribute value obtains modules in the fault detector and fault detector
Attribute value, wherein the attribute value configuration node sample set of modules in the fault detector and fault detector;
3) according to the node sample set, modules are calculated in the fault detector in history working
Under information gain-ratio;
4) it is selected most in the information gain-ratio under history working from modules in the fault detector
Big information gain-ratio, and using the corresponding module of maximum information gain-ratio as current layer root node;
5) using the history working of the corresponding module of the current layer root node as the current layer root node
Branch, and by the branch of current layer root node, the corresponding history of information gain-ratio maximum value of history working
Main split of the working as current layer root node;
6) using other information ratio of profit increase as remaining information ratio of profit increase, the remaining information ratio of profit increase if it exists, from described
Select maximum information gain-ratio in remaining information ratio of profit increase, and by maximum information gain-ratio in the remaining information ratio of profit increase
Leaf node of the corresponding module as the main split of the current layer root node, using the leaf node as new current layer
Root node, and return to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
7) decision tree of the history working of fault detector modules is generated, the decision tree is as failure
The judgment rule of the work at present operating status of indicator;
8) the work at present operating status for acquiring fault detector modules in real time, inputs the decision tree, obtains event
Hinder the diagnostic result of indicator.
With reference to first aspect, in one implementation, the calculation method of the information gain-ratio includes:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set,
The comentropy of the attribute value node sample set of the working of H (D) expression fault detector, and H (D | Ai) indicate failure
The comentropy of the attribute value of the working of each module of indicator, D indicate sample data sets, | D | indicate sample size,
CKIndicate the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and transports
The sum that row state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate section
The set of module i in point sample set, | Di| indicate sample DiNumber, AiIndicate the history work of the module i of fault detector
Operating status set,It is the history working classification according to the module i of fault detector to node sample set D
The K class divided,Indicate the history working classification of the module i of fault detector to node sample set
The sum for the K class that D is divided, n indicate the number of the subset of set D.
With reference to first aspect, in one implementation, the acquisition fault detector and fault detector each module
After history working, further includes:
By the history working of each module of history working and fault detector of fault detector into
Row association, forms the corresponding relationship of working and attribute value.
With reference to first aspect, in one implementation, the input decision tree, obtains the diagnosis of fault detector
As a result, comprising:
According to the work at present operating status for branching into fault detector module of decision route least significant end, determine that failure refers to
Show the work at present operating status of device.
Second aspect, the embodiment of the present application provide a kind of fault diagnosis system of distribution network failure indicator, the event
Hindering diagnostic system includes:
Attribute value setting unit, for be arranged modules in fault detector and fault detector respectively various works
Make the corresponding relationship of the attribute value under operating status;
History working acquisition unit, for acquiring modules in the fault detector and fault detector
History working obtain the failure according to the corresponding relationship of the various workings and attribute value and refer to
Show the attribute value of modules in device and fault detector, wherein modules in the fault detector and fault detector
Attribute value configuration node sample set;
Information gain-ratio computing unit, for calculating each in the fault detector according to the node sample set
Information gain-ratio of the module under history working;
Current layer root node generation unit, for from the fault detector modules in history working
Under information gain-ratio in, select maximum information gain-ratio, and using the corresponding module of maximum information gain-ratio as current
Layer root node;
Main split's generation unit of current layer root node, for the history work of the corresponding module of the current layer root node
Make branch of the operating status as the current layer root node, and by the branch of current layer root node, history work runs shape
Main split of the corresponding history working of the information gain-ratio maximum value of state as current layer root node;
New current layer root node generation unit is used for using other information ratio of profit increase as remaining information ratio of profit increase, if depositing
In the remaining information ratio of profit increase, maximum information gain-ratio is selected from the remaining information ratio of profit increase, and by the residue
The leaf node of main split of the corresponding module of maximum information gain-ratio as the current layer root node in information gain-ratio,
Using the leaf node as new current layer root node, and return to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
Decision tree generation unit, the decision tree of the history working for generating fault detector modules,
Judgment rule of the decision tree as the work at present operating status of fault detector;
Fault detector diagnosis unit, for acquiring the work at present operating status of fault detector modules in real time,
The decision tree is inputted, the diagnostic result of fault detector is obtained.
In conjunction with second aspect, in one implementation, the computing unit of the information gain-ratio, comprising:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set,
The comentropy of the attribute value node sample set of the working of H (D) expression fault detector, and H (D | Ai) indicate failure
The comentropy of the attribute value of the working of each module of indicator, D indicate sample data sets, | D | indicate sample size,
CKIndicate the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and transports
The sum that row state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate section
The set of module i in point sample set, | Di| indicate sample DiNumber, AiIndicate the history work of the module i of fault detector
Operating status set,It is the history working classification according to the module i of fault detector to node sample set D
The K class divided,Indicate the history working classification of the module i of fault detector to node sample set
The sum for the K class that D is divided, n indicate the number of the subset of set D.
In conjunction with second aspect, in one implementation, after the history working acquisition unit, also wrap
It includes:
Associative cell, for by the history work of each module of history working and fault detector of fault detector
It is associated as operating status, forms the corresponding relationship of working and attribute value.
In conjunction with second aspect, in one implementation, the fault detector diagnosis unit, further includes:
Work at present operating status determines subelement, for branching into fault detector mould according to decision route least significant end
The work at present operating status of block determines the work at present operating status of fault detector.
From the above technical scheme, the fault diagnosis side of a kind of distribution network failure indicator provided by the embodiments of the present application
Method and system.The method comprise the steps that setting fault detector and fault detector in modules respectively various works
Make the corresponding relationship of the attribute value under operating status;According to the corresponding relationship of the various workings and attribute value, obtain
Take the attribute value of the history working of modules in the fault detector and fault detector, configuration node sample
Set;According to node sample set, information gain of the modules under history working in fault detector is calculated
Rate;From modules in fault detector in the information gain-ratio under history working, selects maximum information and increase
The corresponding module of beneficial rate is as current layer root node;And it chooses history working in the branch of current layer root node
Main split of the information gain-ratio maximum value as current layer root node;Using other information ratio of profit increase as remaining information ratio of profit increase,
The remaining information ratio of profit increase if it exists, using the corresponding module of information gain-ratio maximum in the remaining information ratio of profit increase as
The leaf node of the main split of the current layer root node using leaf node as new current layer root node, and is selected new
The main split of current layer root node;The remaining information ratio of profit increase if it does not exist, then generate decision tree;It is each to generate fault detector
The decision tree of the history working of a module;The work at present of acquisition fault detector modules runs shape in real time
State obtains the diagnostic result of fault detector according to the decision tree.
When using preceding method or device to be able to solve to solve fault detector itself fail in the prior art, nothing
Method finds the failure of fault detector in time, and can not determine the specific module that indicating fault breaks down, and leads to maintenance workers
The larger problem of journey amount, so that the failure of fault detector is found in time, according to evidence in fault detector itself fail
The branch of decision tree can determine the specific module that indicating fault breaks down, and reduce maintenance engineering amount.
Detailed description of the invention
In order to illustrate more clearly of the technical solution of the application, letter will be made to attached drawing needed in the embodiment below
Singly introduce, it should be apparent that, for those of ordinary skills, without any creative labor,
It is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of flow diagram of the method for diagnosing faults of distribution network failure indicator provided by the present application;
Fig. 2 is a kind of structural representation of the fault diagnosis system of distribution network failure indicator in the application one embodiment
Figure.
Specific embodiment
In order to make the above objects, features, and advantages of the present application more apparent, with reference to the accompanying drawing and it is specific real
Applying mode, the present application will be further described in detail.
The fault diagnosis method and system of a kind of distribution network failure indicator provided by the present application, to solve in the prior art
When fault detector itself fail, the failure of fault detector can not be found in time, and can not determine that indicating fault is sent out
The specific module of raw failure, the problem for causing maintenance engineering amount larger.
Referring to Fig.1, a kind of flow diagram of the method for diagnosing faults of distribution network failure indicator is shown, is specifically included
Following steps:
Step 101, be arranged modules in fault detector and fault detector respectively various workings
Under attribute value corresponding relationship;
Fault detector for distribution network failure positioning is mainly by fault detector ontology, communication terminal module and signal
Source module composition, the major functions of each module are as follows:
Fault detector ontology is flowed through for detecting the presence of fault current, and when the module failure, fault detector be will be unable to
It works normally;
Communication module, the fault current information for detecting fault detector ontology pass control main website back, the module
When breaking down, fault detector will be unable to work normally;
Signal source module, for the Injection Signal into the power grid that singlephase earth fault occurs, support fault detector detection
And the singlephase earth fault in power grid is positioned, when which breaks down, fault detector will be unable to detection singlephase earth fault;
The working of fault detector refers to overall operation situation, as fault detector normal operation, equipment are stopped
With, equipment fault etc.;
The module working of fault detector refers to the practival operating condition of modules in fault detector, such as
Main body, communication terminal, signal source of fault detector etc.;
In this step, each working of fault locator and each module of fault locator is drawn according to function
Point, corresponding attribute value successively encodes;
Such as: when the module failure of fault detector, fault detector is caused to be unable to normal use, equipment running status
Attribute value is set as 2- failure, and when fault detector ontology breaks down, the working of fault detector is can not to transport
Row, as breaks down;
When the working of fault detector module is alarm, then fault detector operating status is set as 1- announcement
It is alert;
When the working of all modules of fault detector is normal, fault detector operating status is set as
0- is normal;
Below by taking the communication terminal module of fault detector as an example, communication terminal module includes: data transmission, data storage
It deposits, clock clock synchronization and host power supply, referring to table 1, the communication terminal module working and indicating fault of fault detector
The attribute value of the working of device;
Table 1: the working of the communication terminal module working and fault detector of fault detector
Attribute value
As can be seen from Table 1: the data transfer characteristics value of communication terminal module is 0- normal transmission, corresponding fault detector
Working attribute value be 0- it is normal;The data transfer characteristics value of communication terminal module is that 1- can not be transmitted, corresponding event
The working attribute value for hindering indicator is 2- failure;
The data storage attribute value of communication terminal module is 0- normal storage, the working of corresponding fault detector
Attribute value is that 0- is normal;The data storage attribute value of communication terminal module is that 1- storage volume has expired, corresponding fault detector
Working attribute value is 1- alarm;
The clock clock synchronization attribute of communication terminal module is the normal clock synchronization of 0-, the working category of corresponding fault detector
Property value be 0- it is normal;The clock clock synchronization attribute of communication terminal module be 1- can not clock synchronization, the work operation of corresponding fault detector
State attribute value is 2- failure;
The host power supply attribute value of communication terminal module is that 0- power supply is normal, the working of corresponding fault detector
Attribute value is that 0- is normal;The host power supply attribute value of communication terminal module is that 1- cell host is under-voltage, corresponding fault detector fortune
Row Status Level 1- alarm.
The setting of the attribute value of other modules of fault detector and can be with reference to logical with the corresponding relationship of fault detector
Believe the setting method of the attribute value of terminal module.
Step 102, the history working of modules in the fault detector and fault detector, root are acquired
According to the corresponding relationship of the various workings and attribute value, obtain each in the fault detector and fault detector
The attribute value of module, wherein the attribute value configuration node sample set of modules in the fault detector and fault detector
It closes;
In this step, by acquiring the history working of modules in fault detector and fault detector,
The corresponding relationship of the working and attribute value that are arranged according to step 101, by attribute value configuration node sample set.
By taking communication terminal as an example: referring to table 2, the working historical data of fault detector communication terminal module with
And the node sample set that the working attribute value of corresponding fault detector is formed.
The node sample set of 2 communication terminal module of table
Step 103, according to the node sample set, modules in the fault detector is calculated and are transported in history work
Information gain-ratio under row state;
In this step, according to the node sample set formed in step 102, each mould in fault detector is calculated separately
Information gain-ratio under the history working of block, i.e., each corresponding information gain-ratio of module.
Step 104, from modules in the fault detector in the information gain-ratio under history working,
Maximum information gain-ratio is selected, and using the corresponding module of maximum information gain-ratio as current layer root node;
In this step, from selected in step 103 the corresponding module of information gain-ratio maximum value as current layer root node.
Step 105, using the history working of the corresponding module of the current layer root node as the current layer
The branch of root node, and by the branch of current layer root node, the information gain-ratio maximum value of history working is corresponding
Main split of the history working as current layer root node;
In this step, the history working of the corresponding module of front layer root node is as the current layer root node
Branch, because the history working of module has multiple, then branch has multiple, therefore selects the letter of history working
Cease main split of the corresponding history working of ratio of profit increase maximum value as current layer root node.
Step 106, it using other information ratio of profit increase as remaining information ratio of profit increase, judges whether there is the remaining information and increases
Beneficial rate;
Step 107, the remaining information ratio of profit increase if it exists, selects maximum information from the remaining information ratio of profit increase
Ratio of profit increase, and using the corresponding module of information gain-ratio maximum in the remaining information ratio of profit increase as the current layer root node
Main split leaf node, using the leaf node as new current layer root node, and return to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
In this step, using the other information ratio of profit increase in step 104 as remaining information ratio of profit increase, the residue if it exists
Information gain-ratio, selects maximum information gain-ratio from the remaining information ratio of profit increase, and by the remaining information ratio of profit increase
In leaf node of the corresponding module of maximum information gain-ratio as the main split of current layer root node in step 105, i.e. shape
The root node of the current layer of Cheng Xin, and from the history working of the new corresponding module of front layer root node as described new
Current layer root node main split;If there is no remaining information ratio of profit increase, then the operation of step 7) is directly executed.
Step 108, the decision tree of the history working of fault detector modules is generated, the decision tree makees
For the judgment rule of the work at present operating status of fault detector;
In this step, by step 101 to 106, the history working of fault detector modules is formd
Decision tree, using the decision tree, to diagnose fault the failure of indicator.
Step 109, the work at present operating status for acquiring fault detector modules in real time, inputs the decision tree,
Obtain the diagnostic result of fault detector.
In this step, the decision tree generated from step 108 is analyzed, and judgment rule is generated, to diagnose fault indicator
Working, to judge the failure of fault detector;
By taking communication terminal module as an example, the judgment rule are as follows: IF communication terminal data transmission working be
The working of " 0- normal transmission " AND clock clock synchronization is the working of " the normal clock synchronization of 0- " AND data storage
Working for " 0- normal storage " THEN fault detector is " 0- is normal ";
The working of IF communication terminal data transmission is that the work of " 0- normal transmission " AND clock clock synchronization runs shape
State is that the working of " the normal clock synchronization of 0- " AND data storage is the work of " 1- storage volume has expired " THEN fault detector
Making operating status is " 1- alarm ";
The working of IF communication terminal data transmission is that the work of " 0- normal transmission " AND clock clock synchronization runs shape
State is that the working of " 1- can not clock synchronization " THEN fault detector is " 2- failure ";
The working of IF communication terminal data transmission is that the work of " 1- can not be transmitted " THEN fault detector is transported
Row state is " 2- failure ".
The method for diagnosing faults of a kind of distribution network failure indicator provided by the embodiments of the present application, comprising: setting failure refers to
Show modules in device and fault detector respectively various workings under attribute value corresponding relationship;According to institute
The corresponding relationship for stating various workings and attribute value obtains modules in the fault detector and fault detector
History working attribute value, configuration node sample set;According to node sample set, calculate in fault detector
Information gain-ratio of the modules under history working;Modules work in history and run from fault detector
In information gain-ratio under state, the corresponding module of maximum information gain-ratio is selected as current layer root node;And choose by
Master point of the information gain-ratio maximum value of history working as current layer root node in the branch of current layer root node
Branch;Using other information ratio of profit increase as remaining information ratio of profit increase, the remaining information ratio of profit increase if it exists, by the remaining information
The leaf node of main split of the corresponding module of maximum information gain-ratio as the current layer root node in ratio of profit increase, by leaf
Child node selects the main split of new current layer root node as new current layer root node;The remaining letter if it does not exist
Ratio of profit increase is ceased, then generates decision tree;Generate the decision tree of the history working of fault detector modules;It adopts in real time
The work at present operating status of collection fault detector modules obtains the diagnosis knot of fault detector according to the decision tree
Fruit.
When using preceding method or device to be able to solve to solve fault detector itself fail in the prior art, nothing
Method finds the failure of fault detector in time, and can not determine the specific module that indicating fault breaks down, and leads to maintenance workers
The larger problem of journey amount, so that the failure of fault detector is found in time, according to evidence in fault detector itself fail
The branch of decision tree can determine the specific module that indicating fault breaks down, and reduce maintenance engineering amount.
Further, the calculation method of the information gain-ratio includes:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set,
The comentropy of the attribute value node sample set of the working of H (D) expression fault detector, and H (D | Ai) indicate failure
The comentropy of the attribute value of the working of each module of indicator, D indicate sample data sets, | D | indicate sample size,
CKIndicate the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and transports
The sum that row state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate section
The set of module i in point sample set, | Di| indicate sample DiNumber, AiIndicate the history work of the module i of fault detector
Operating status set,It is the history working classification according to the module i of fault detector to node sample set D
The K class divided,Indicate the history working classification of the module i of fault detector to node sample set
The sum for the K class that D is divided, n indicate the number of the subset of set D.
In the present embodiment, according to node sample set D, this sample data of communication terminal module training set data is divided into 3 classes,
The operating status 0- of fault detector is normal, 1- alarm, 2- failure;
Communication terminal module training set has 4 module attributes, module attribute set={ data transmission, data preservation, clock
Clock synchronization, host power supply }.
The calculating of information gain-ratio of the communication terminal module under history working in the fault detector
Journey is as follows:
1. calculate the comentropy H (D) in sample set, in sample the working of fault detector have 3 it is normal,
4 alarms, 5 failures, according to formula:
H (D)=- 3/12*log (3/12) -4/12*log (4/12) -5/12*log (5/12)
2. calculate each module all working operating status of communication terminal comentropy H (D | Ai), with communication terminal module
For, it is 0 for communication terminal data transmission working attribute value, a total of 8, wherein fault indicator is transported
Row state has 3 normal, 4 alarms, 1 failure;Working attribute value 1 is transmitted for communication terminal data, in total
There are 4, wherein fault indicator operating status has 0 normal, 0 alarm, 4 failures, according to formulaThe comentropy of fault detector communication terminal data transfer characteristics value are as follows:
H(D|Ai)=8/12* (- 3/8*log (3/8) -4/8*log (4/8) -1/8*log (1/8) -4/12* (- 0-0-4/
4*log(4/4))
Similarly, communication terminal data storage, clock clock synchronization, the comentropy calculation method of host power supply attribute value and communicate eventually
The calculation method of the comentropy of end data transmission property value is identical.
3. the comentropy of all working operating status attribute value of module in calculate node sample setFor number
According to transmission, the attribute value of working is a total of 8 of 0, and attribute value is a total of 4 of 1, according to formulaCalculate the comentropy of the attribute value of the working of data transmission are as follows:
Similarly, the comentropy calculating side of the attribute value of the working of data storage, clock clock synchronization and host power supply
Method is identical as the calculation method of the comentropy of the attribute value for the working that data are transmitted.
4. according to g (D, Ai)=H (D)-H (D | Ai), information gain g (D, A is calculatedi)。
5. basisInformation gain-ratio is calculated.
Further, after the history working for acquiring fault detector and each module of fault detector,
Further include:
By the history working of each module of history working and fault detector of fault detector into
Row association, forms the corresponding relationship of working and attribute value.
Wherein, associated to be meant that every kind of working module in fault detector corresponds to fault detector
Certain a kind of working, specific correlating method is referring to described in step 101;
Further, the input decision tree, obtains the diagnostic result of fault detector, comprising:
According to the work at present operating status for branching into fault detector module of decision route least significant end, determine that failure refers to
Show the work at present operating status of device.
In a kind of method for diagnosing faults of distribution network failure indicator of the application, according to correlation rule, can directly from
The working of module is inferred to the working of fault detector.
Referring to Fig. 2, a kind of structural schematic diagram of the fault diagnosis system of distribution network failure indicator is shown, is specifically included
Such as flowering structure:
Attribute value setting unit 100, for be arranged modules in fault detector and fault detector respectively it is each
The corresponding relationship of attribute value under kind working;
History working acquisition unit 200, it is each in the fault detector and fault detector for acquiring
The history working of module obtains the event according to the corresponding relationship of the various workings and attribute value
Hinder the attribute value of modules in indicator and fault detector, wherein each in the fault detector and fault detector
The attribute value configuration node sample set of module;
Information gain-ratio computing unit 300, it is each in the fault detector for calculating according to the node sample set
Information gain-ratio of a module under history working;
Current layer root node generation unit 400 is used for the modules from the fault detector and runs in history work
In information gain-ratio under state, select maximum information gain-ratio, and using the corresponding module of maximum information gain-ratio as
Current layer root node;
Main split's generation unit 500 of current layer root node, for going through with the corresponding module of current layer root node
Branch of the history working as the current layer root node, and by the branch of current layer root node, history work is transported
Main split of the corresponding history working of the information gain-ratio maximum value of row state as current layer root node;
Judging unit 600 judges whether there is described for sentencing using other information ratio of profit increase as remaining information ratio of profit increase
Remaining information ratio of profit increase;
New current layer root node generation unit 700 is used for using other information ratio of profit increase as remaining information ratio of profit increase, if
There are the remaining information ratio of profit increase, maximum information gain-ratio is selected from the remaining information ratio of profit increase, and will be described surplus
The leaf section of main split of the corresponding module of maximum information gain-ratio as the current layer root node in remaining information gain-ratio
Point using the leaf node as new current layer root node, and returns to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
Decision tree generation unit 800, the decision of the history working for generating fault detector modules
Tree, judgment rule of the decision tree as the work at present operating status of fault detector;
Fault detector diagnosis unit 900, the work at present for acquiring fault detector modules in real time run shape
State inputs the decision tree, obtains the diagnostic result of fault detector.
Further, the computing unit of the information gain-ratio, comprising:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set,
The comentropy of the attribute value node sample set of the working of H (D) expression fault detector, and H (D | Ai) indicate failure
The comentropy of the attribute value of the working of each module of indicator, D indicate sample data sets, | D | indicate sample size,
CKIndicate the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and transports
The sum that row state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate section
The set of module i in point sample set, | Di| indicate sample DiNumber, AiIndicate the history work of the module i of fault detector
Operating status set,It is the history working classification according to the module i of fault detector to node sample set D
The K class divided,Indicate the history working classification of the module i of fault detector to node sample set
The sum for the K class that D is divided, n indicate the number of the subset of set D.
Further, after the history working acquisition unit, further includes:
Associative cell, for by the history work of each module of history working and fault detector of fault detector
It is associated as operating status, forms the corresponding relationship of working and attribute value.
Further, the fault detector diagnosis unit, further includes:
Work at present operating status determines subelement, for branching into fault detector mould according to decision route least significant end
The work at present operating status of block determines the work at present operating status of fault detector.
In the specific implementation, the application also provides a kind of computer storage medium, wherein the computer storage medium can store
There is program, which may include a kind of each reality of the method for diagnosing faults of distribution network failure indicator provided by the present application when executing
Some or all of apply in example step.The storage medium can be magnetic disk, CD, read-only memory (English: read-
Only memory, referred to as: ROM) or random access memory (English: random access memory, referred to as: RAM) etc..
It is required that those skilled in the art can be understood that the technology in the embodiment of the present application can add by software
The mode of general hardware platform realize.Based on this understanding, the technical solution in the embodiment of the present application substantially or
Say that the part that contributes to existing technology can be embodied in the form of software products, which can deposit
Storage is in storage medium, such as ROM/RAM, magnetic disk, CD, including some instructions are used so that computer equipment (can be with
It is personal computer, server or the network equipment etc.) execute certain part institutes of each embodiment of the application or embodiment
The method stated.
Same and similar part may refer to each other between each embodiment in this specification.Implement especially for device
For example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring in embodiment of the method
Explanation.
Above-described the application embodiment does not constitute the restriction to the application protection scope.
Claims (8)
1. a kind of method for diagnosing faults of distribution network failure indicator characterized by comprising
1) be arranged modules in fault detector and fault detector respectively various workings under attribute value
Corresponding relationship;
2) the history working for acquiring modules in the fault detector and fault detector, according to described various
The corresponding relationship of working and attribute value obtains the attribute of modules in the fault detector and fault detector
Value, wherein the attribute value configuration node sample set of modules in the fault detector and fault detector;
3) according to the node sample set, modules are calculated in the fault detector under history working
Information gain-ratio;
4) it is selected maximum in the information gain-ratio under history working from modules in the fault detector
Information gain-ratio, and using the corresponding module of maximum information gain-ratio as current layer root node;
5) point using the history working of the corresponding module of the current layer root node as the current layer root node
Branch, and by the branch of current layer root node, the corresponding history of the information gain-ratio maximum value of history working works
Main split of the operating status as current layer root node;
6) using other information ratio of profit increase as remaining information ratio of profit increase, the remaining information ratio of profit increase if it exists, from the residue
Maximum information gain-ratio is selected in information gain-ratio, and information gain-ratio maximum in the remaining information ratio of profit increase is corresponding
Module as the current layer root node main split leaf node, using the leaf node as new current layer root section
Point, and return to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
7) decision tree of the history working of fault detector modules is generated, the decision tree is as indicating fault
The judgment rule of the work at present operating status of device;
8) the work at present operating status for acquiring fault detector modules in real time, inputs the decision tree, obtains failure and refer to
Show the diagnostic result of device.
2. the method for diagnosing faults of distribution network failure indicator according to claim 1, which is characterized in that the information increases
The calculation method of beneficial rate includes:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set, H (D)
The comentropy of the attribute value node sample set of the working of expression fault detector, and H (D | Ai) indicate indicating fault
The comentropy of the attribute value of the working of each module of device, D indicate sample data sets, | D | indicate sample size, CKTable
Show the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and runs
The sum that state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate node
The set of module i in sample set, | Di| indicate sample DiNumber, AiIt indicates that the history of the module i of fault detector works to transport
Row state set,Be according to the module i of fault detector history working classification to node sample set D into
The K class that row divides,Indicate fault detector module i history working classification to node sample set D into
The sum for the K class that row divides, n indicate the number of the subset of set D.
3. the method for diagnosing faults of distribution network failure indicator according to claim 1, which is characterized in that the acquisition event
After barrier indicator and the history working of each module of fault detector, further includes:
The history working of each module of history working and fault detector of fault detector is closed
Connection forms the corresponding relationship of working and attribute value.
4. the method for diagnosing faults of distribution network failure indicator according to claim 1, which is characterized in that the input institute
Decision tree is stated, the diagnostic result of fault detector is obtained, comprising:
According to the work at present operating status for branching into fault detector module of decision route least significant end, fault detector is determined
Work at present operating status.
5. a kind of fault diagnosis system of distribution network failure indicator characterized by comprising
Attribute value setting unit, for be arranged modules in fault detector and fault detector respectively various work fortune
The corresponding relationship of attribute value under row state;
History working acquisition unit, for acquiring going through for modules in the fault detector and fault detector
History working obtains the fault detector according to the corresponding relationship of the various workings and attribute value
With the attribute value of modules in fault detector, wherein the category of modules in the fault detector and fault detector
Property value configuration node sample set;
Information gain-ratio computing unit, for calculating modules in the fault detector according to the node sample set
Information gain-ratio under history working;
Current layer root node generation unit, for from modules in the fault detector under history working
In information gain-ratio, maximum information gain-ratio is selected, and using the corresponding module of maximum information gain-ratio as current layer root
Node;
Main split's generation unit of current layer root node is transported for being worked with the history of the corresponding module of the current layer root node
Branch of the row state as the current layer root node, and by the branch of current layer root node, history working
Main split of the corresponding history working of information gain-ratio maximum value as current layer root node;
New current layer root node generation unit, for using other information ratio of profit increase as remaining information ratio of profit increase, institute if it exists
Remaining information ratio of profit increase is stated, selects maximum information gain-ratio from the remaining information ratio of profit increase, and by the remaining information
The leaf node of main split of the corresponding module of maximum information gain-ratio as the current layer root node in ratio of profit increase, by institute
Leaf node is stated as new current layer root node, and returns to step operation 5);
The remaining information ratio of profit increase if it does not exist executes the operation of step 7);
Decision tree generation unit, the decision tree of the history working for generating fault detector modules are described
Judgment rule of the decision tree as the work at present operating status of fault detector;
Fault detector diagnosis unit is inputted for acquiring the work at present operating status of fault detector modules in real time
The decision tree obtains the diagnostic result of fault detector.
6. the fault diagnosis system of distribution network failure indicator according to claim 5, which is characterized in that the information increases
The computing unit of beneficial rate, comprising:
G (D, Ai)=H (D)-H (D | Ai)
Wherein, g (D, Ai) indicate information gain,Indicate the comentropy of feature A in the node sample set, H (D)
The comentropy of the attribute value node sample set of the working of expression fault detector, and H (D | Ai) indicate indicating fault
The comentropy of the attribute value of the working of each module of device, D indicate sample data sets, | D | indicate sample size, CKTable
Show the set of the working of fault detector, | CK| it indicates that the history of the K class state of fault detector works and runs
The sum that state occurs in node sample, K indicate the classification of the history working of fault detector, DiIndicate node
The set of module i in sample set, | Di| indicate sample DiNumber, AiIt indicates that the history of the module i of fault detector works to transport
Row state set,Be according to the module i of fault detector history working classification to node sample set D into
The K class that row divides,Indicate fault detector module i history working classification to node sample set D into
The sum for the K class that row divides, n indicate the number of the subset of set D.
7. the fault diagnosis system of distribution network failure indicator according to claim 5, which is characterized in that the history work
After making operating status acquisition unit, further includes:
Associative cell is transported for the history of each module of history working and fault detector of fault detector to work
Row state is associated, and forms the corresponding relationship of working and attribute value.
8. the fault diagnosis system of distribution network failure indicator according to claim 5, which is characterized in that the failure refers to
Show device diagnosis unit, further includes:
Work at present operating status determines subelement, for branching into fault detector module according to decision route least significant end
Work at present operating status determines the work at present operating status of fault detector.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811001810.3A CN109188330A (en) | 2018-08-30 | 2018-08-30 | A kind of fault diagnosis method and system of distribution network failure indicator |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811001810.3A CN109188330A (en) | 2018-08-30 | 2018-08-30 | A kind of fault diagnosis method and system of distribution network failure indicator |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109188330A true CN109188330A (en) | 2019-01-11 |
Family
ID=64917147
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811001810.3A Pending CN109188330A (en) | 2018-08-30 | 2018-08-30 | A kind of fault diagnosis method and system of distribution network failure indicator |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109188330A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113051553A (en) * | 2019-12-27 | 2021-06-29 | 北京安控科技股份有限公司 | Accident rule generation method and system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7349889B1 (en) * | 2000-11-20 | 2008-03-25 | Rohm And Haas Electronic Materials Llc | System and method for remotely diagnosing faults |
CN104111433A (en) * | 2014-07-16 | 2014-10-22 | 云南电力试验研究院(集团)有限公司电力研究院 | Device for detecting motion performances of earth fault indicator |
CN203965613U (en) * | 2014-06-13 | 2014-11-26 | 云南电力试验研究院(集团)有限公司电力研究院 | A kind of emulation test system of earth-fault indicator performance |
CN104483643A (en) * | 2014-11-27 | 2015-04-01 | 华中科技大学 | Modeling method based on detection platform of fault indicator for physical simulation of power distribution network |
CN105894177A (en) * | 2016-03-25 | 2016-08-24 | 国家电网公司 | Decision-making-tree-algorithm-based analysis and evaluation method for operation risk of power equipment |
CN106600163A (en) * | 2016-12-26 | 2017-04-26 | 中电长城(长沙)信息技术有限公司 | Financial self-service terminal fault diagnosis method based on decision tree learning algorithm and system thereof |
CN106934423A (en) * | 2017-03-16 | 2017-07-07 | 重庆邮电大学 | The construction method and system of a kind of decision tree |
-
2018
- 2018-08-30 CN CN201811001810.3A patent/CN109188330A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7349889B1 (en) * | 2000-11-20 | 2008-03-25 | Rohm And Haas Electronic Materials Llc | System and method for remotely diagnosing faults |
CN203965613U (en) * | 2014-06-13 | 2014-11-26 | 云南电力试验研究院(集团)有限公司电力研究院 | A kind of emulation test system of earth-fault indicator performance |
CN104111433A (en) * | 2014-07-16 | 2014-10-22 | 云南电力试验研究院(集团)有限公司电力研究院 | Device for detecting motion performances of earth fault indicator |
CN104483643A (en) * | 2014-11-27 | 2015-04-01 | 华中科技大学 | Modeling method based on detection platform of fault indicator for physical simulation of power distribution network |
CN105894177A (en) * | 2016-03-25 | 2016-08-24 | 国家电网公司 | Decision-making-tree-algorithm-based analysis and evaluation method for operation risk of power equipment |
CN106600163A (en) * | 2016-12-26 | 2017-04-26 | 中电长城(长沙)信息技术有限公司 | Financial self-service terminal fault diagnosis method based on decision tree learning algorithm and system thereof |
CN106934423A (en) * | 2017-03-16 | 2017-07-07 | 重庆邮电大学 | The construction method and system of a kind of decision tree |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113051553A (en) * | 2019-12-27 | 2021-06-29 | 北京安控科技股份有限公司 | Accident rule generation method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104020754B (en) | Method for enabling state monitoring information of transformer station primary main equipment to access to regulation and control system | |
CN109948808A (en) | The banking process in substation equipment fault case library, fault diagnosis method and system | |
CN104796273A (en) | Method and device for diagnosing root of network faults | |
CN112415331B (en) | Power grid secondary system fault diagnosis method based on multi-source fault information | |
CN102055187B (en) | Large-scale interconnected power grid spinning reserve risk assessment method based on state space division method | |
CN107633307B (en) | Power supply and distribution system root alarm detection method, device, terminal and computer storage medium | |
CN105894177A (en) | Decision-making-tree-algorithm-based analysis and evaluation method for operation risk of power equipment | |
CN107069960A (en) | A kind of online defect diagnostic method of secondary operation management system | |
CN105548744A (en) | Substation equipment fault identification method based on operation-detection large data and system thereof | |
CN111143438A (en) | Workshop field data real-time monitoring and anomaly detection method based on stream processing | |
CN108616405A (en) | The Fault Locating Method and device in a kind of intelligent substation void circuit | |
CN105719062A (en) | Method for assessing risks and weak links of power grid, with double fault probability characteristics considered | |
CN116049146B (en) | Database fault processing method, device, equipment and storage medium | |
CN106603274A (en) | Power distribution network fault locating method based on multidimensional communication data | |
CN107453354A (en) | A kind of weak link recognition methods of power distribution network | |
CN110311802A (en) | Network operation method, device, electronic equipment and storage medium | |
CN106296440A (en) | Based on transformer station's warning information analysis and decision system integrated for ANN and ES and method | |
CN113740666B (en) | Method for positioning root fault of storm alarm in power system of data center | |
CN109188330A (en) | A kind of fault diagnosis method and system of distribution network failure indicator | |
CN114676791A (en) | Electric power system alarm information processing method based on fuzzy evidence reasoning | |
CN111027721B (en) | System fault positioning method | |
CN109639475A (en) | Network self-diagnosis Fault Locating Method based on associated diagram | |
CN117034149A (en) | Fault processing strategy determining method and device, electronic equipment and storage medium | |
CN108847955A (en) | A kind of fault pre-alarming monitoring method of equipment and business | |
CN110021933B (en) | Power information system control function reliability assessment method considering component faults |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190111 |