CN106405352A - Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant - Google Patents
Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant Download PDFInfo
- Publication number
- CN106405352A CN106405352A CN201611005524.5A CN201611005524A CN106405352A CN 106405352 A CN106405352 A CN 106405352A CN 201611005524 A CN201611005524 A CN 201611005524A CN 106405352 A CN106405352 A CN 106405352A
- Authority
- CN
- China
- Prior art keywords
- esdd
- early warning
- insulator
- unit
- neural network
- 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
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
- G01R31/1245—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of line insulators or spacers, e.g. ceramic overhead line cap insulators; of insulators in HV bushings
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/185—Electrical failure alarms
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Chemical & Material Sciences (AREA)
- Engineering & Computer Science (AREA)
- Ceramic Engineering (AREA)
- Insulators (AREA)
Abstract
The invention discloses an equivalent salt deposit density (ESDD) prediction and early warning system for a power insulator surface contaminant, which comprises four function units of a genetic algorithm training unit, a BP (Back Propagation) neural network training unit, a genetic BP neural network predication unit and an early warning unit. The genetic algorithm is good at global search, the BP neural network is good at local search, and thus, when the genetic algorithm and the BP algorithm are combined, the prediction accuracy can be improved, and the algorithm convergence rate is improved. The genetic algorithm firstly optimizes the initial weight of the neural network, good search space is positioned, and the BP algorithm is then adopted to search the optimal value in the small space. The application conditions are wide, the system can be applied to insulators of any types and is not influenced by different contaminant accommodation features due to different insulator types, and the prediction method and the early warning system can realize insulator ESDD prediction, and give contaminant flashover and early warning signals.
Description
Technical field
The present invention relates to pollution severity of insulators early warning technology field, specially a kind of electric insulator surface area dirt
Equivalent salt density prediction and early warning system.
Background technology
Insulator under normal working voltage, due to the accumulation of surface filth thing, in the boisterous work such as wet weather, dense fog
With under be susceptible to pollution flashover accident, and pollution flashover accident can constitute serious prestige to the safe and stable operation of power system
The side of body.Therefore, the pollution degree of insulator on transmission line of electricity is predicted being highly desirable to, so that the sending out of timely Prevent from Dirt Flash accident
Raw.Generally, people use equivalent salt deposit density (Equivalent Salt Deposit Density, ESDD), referred to as equivalent salt
Close assessing pollution severity of insulators degree.
In recent years, neutral net because its have MPP ability, strong fault tolerance, self-organizing and adaptive should be able to
Power is strong and the features such as association function is strong, it has also become the powerful of solve problem, is widely used in many scientific domains.
But, because it also exists, convergence rate is slow, is easily trapped into the defects such as local minimum points so that the prediction of general neural network is smart
Degree is not high.It is therefore proposed that various algorithms to be optimized process to general neural network.
Back propagation (Back Propagation, BP) is as long as neutral net has enough hidden layers and hidden node, just
Nonlinear mapping relation can be approached with arbitrary accuracy, there is preferable learning capacity.Genetic algorithm (Genetic
Algorithm, GA) it is a kind of parallel random search optimization method, there is ability of searching optimum.By genetic algorithm and BP nerve
Network integration can using genetic algorithm from complicated, non-linear and non-differentiable function global search go out preferable solve empty
Between function, search out the work(of optimal solution in these little solution spaces using the None-linear approximation ability of BP neural network simultaneously
Energy.The method can effectively solve genetic algorithm local optimal searching scarce capacity and search space change adaptability difference is lacked
Point, improves the convergence rate of BP neural network simultaneously.
Content of the invention
It is an object of the invention to proposition is a kind of may occur because dunghill accumulation is excessive to insulator on working line
The Forecasting Methodology of pollution flashover accident and early warning system, its applicable elements extensively, can be used on the insulator of disposable type,
Do not affected by because insulator type is different, its contamination accumulation characteristics is different, this Forecasting Methodology and early warning system are capable of to exhausted
Edge ESDD predicts, and sends the function of pre-warning signal.
A kind of equivalent salt density prediction of electric insulator surface area dirt and early warning system, train including including genetic algorithm
Unit, BP neural network training unit, Genetic BP Neutral Network predicting unit, this four functional units of prewarning unit.
Genetic algorithm is good at global search, and BP neural network is good at Local Search, therefore by genetic algorithm and BP algorithm phase
In conjunction with the accuracy that both can improve prediction, improve convergence of algorithm speed.Genetic algorithm is entered to neutral net initial weight first
Row optimizes, and positions search space, then searches for optimal value in little space using BP algorithm.
Genetic BP Neutral Network predicting unit by front once arrive prediction during daily wind speed, precipitation, relative humidity,
The summed data of AQI and front primary insulation ESDD data put into Genetic BP Neutral Network predicting unit.In Genetic BP nerve net
In network predicting unit, the weights of input layer to the hidden layer that note BP neural network training unit obtains are Wij, hidden layer neuron
Threshold value be θj, the connection weight of hidden layer to output layer is Vij, the threshold value of output layer is
The then input of each neuron of hidden layer is
The transmission function of neutral net adopts Sigmoid function f (x)=1/ (1+e-x), then hidden layer is output as
Therefore obtain the input of output layer neuron and be output as
Prewarning unit arranges A, B, C, D, 4 advanced warning grades.Wherein when insulator ESDD numerical value reaches it may happen that pollution flashover
When insulator ESDD numerical value 95% when, i.e. 95% ρF, unit sends A level early warning；When insulator ESDD numerical value reaches and may send out
During raw pollution flashover insulator ESDD numerical value 90% when, i.e. 90% ρF, unit sends B level early warning；When insulator ESDD numerical value reaches
It may happen that during pollution flashover insulator ESDD numerical value 85% when, i.e. 85% ρF, unit sends C level early warning；When insulator ESDD number
Value reach it may happen that during pollution flashover insulator ESDD numerical value 80% when, i.e. 80% ρF, unit sends D level early warning.
BP neural network is a kind of multilayer feedforward neural network.The topological structure of 3 layers of BP network, including input layer, implies
Layer and output layer, each neuron is connected with next layer of all of neuron, with connectionless between layer neuron.The setting of hidden layer
According to Kolmogorov theorem, if output layer has n nodes, the nodes of hidden layer are 2n+1.
A kind of equivalent salt density Forecasting Methodology of electric insulator surface area dirt, comprises the following steps：
1) sample data is put into genetic algorithm functional unit, before sample data, once arrive the every day breeze during prediction
Speed, precipitation, relative humidity, the summed data of AQI and a front ESDD numerical value totally five parameters as input quantity, when secondary
ESDD prediction data, as output, through initialization population, calculates fitness function, selection operation, crossover operation and variation
The process of operation obtains the weights optimizing and threshold value；
2) weights and threshold value that process the optimization obtaining from genetic algorithm training unit are put into BP neural network function list
Unit, as initial weight and threshold value, in BP neural network functional unit, equally once arrived the every day breeze during prediction in the past
Speed, precipitation, relative humidity, the summed data of AQI and a front ESDD numerical value totally five parameters as input quantity, as secondary ESDD
Prediction data, as output, through the Error Calculation of hidden layer and output layer, constantly adjusts weights by training sample training
And threshold value, obtain best weight value and the threshold value of model；
3) best weight value being obtained through sample training according to BP neural network functional unit and threshold value, set up insulator
ESDD forecast model, by the front daily wind speed once arriving during prediction, precipitation, relative humidity, the summed data of AQI and previous
Secondary ESDD numerical value puts into forecast model, output prediction insulator ESDD numerical value；
4) the insulator ESDD numerical value of prediction is put into early warning criterion unit, if ESDD numerical value reaches generation pollution flashover thing
Therefore threshold value then alert, the threshold value of pollution flashover flashover early warning arranges according to regional, different voltages of different gradation for surface pollution etc.
Level and different insulative configuration situations such as and determine.
Genetic algorithm and BP neural network are combined to set up the forecast model of insulator ESDD.
Using weights after the optimization that genetic algorithm exports and threshold value as the initial weight of BP neural network and threshold value, with training
Sample is trained to BP neural network, thus setting up the forecast model of insulator ESDD.By the optimization of genetic algorithm, improve BP god
Convergence rate through network, reduces the probability that BP algorithm is easily absorbed in locally optimal solution, thus realizing ability of searching optimum.
Collecting some cycles inner insulator ESDD, wind speed, precipitation, relative humidity and air quality index (Air
Quality Index, AQI) on the basis of data, set up the BP neural network forecast model of an insulator ESDD.Wherein, defeated
Enter layer and arrange before 5 nodes are respectively, once arrive daily wind speed during predicting, precipitation, relative humidity, the summed data of AQI
With a front ESDD numerical value, it is insulator ESDD predictive value that output layer arranges 1 node, the installation warrants of node in hidden layer
Kolmogorov theorem, is 11 nodes.The initial weight of this BP neural network and threshold value are calculated by genetic algorithm.
In sum, using prediction and the early warning system of insulator dirty degree proposed by the present invention, inhomogeneity can be directed to
The insulator dirty degree of type is effectively predicted.On the basis of obtaining insulator ESDD and meteorological data etc., the present invention passes through
Set up based on genetic algorithm optimization BP neural network to the prediction of insulator and Early-warning Model, can effectively predict insulator
The accumulation of ESDD, sends early warning information in time and sends out arranging insulator cleaning operation reliably to reduce pollution flashover accident
Raw chance.
Brief description
The invention will be further described below in conjunction with the accompanying drawings：
Fig. 1 is prediction and the early warning system flow chart of insulator ESDD；
Fig. 2 is genetic algorithm flow chart；
Fig. 3 is BP neural network structure chart；
Fig. 4 is BP neural network unit flow chart.
Specific embodiment
Below in conjunction with the accompanying drawings 1～4 and embodiment the inventive method is done and clearly and completely describes further, but the present invention
Embodiment be not limited to this.
The present invention proposes a kind of BP neural network based on genetic algorithm optimization to insulator contamination equivalent salt density ESDD's
Prediction and early warning system, Fig. 1 is this working-flow figure, comprises genetic algorithm training unit altogether, and BP neural network training is single
Unit, Genetic BP Neutral Network predicting unit, this four functional units of prewarning unit.
When collecting insulator ESDD historical data, should be noted some requirement following：
(1) consider that it is unpractical for being collected in insulator ESDD data on operation transmission line of electricity daily, the therefore present invention
Carried using the close on-line monitoring equipment of optical sensor power transmission and transformation salt or other on-Line Monitor Device when collecting insulator ESDD data
For data.
(2) when collecting daily meteorological data it is necessary to be the meteorological data around test insulator, it is not suitable for using distance
Meteorological data collected by test insulator place farther out.
To each unit of the present invention be launched to be specifically described below：
Genetic algorithm training unit
As shown in Fig. 2 genetic algorithm is good at global search, BP neural network is good at Local Search, therefore by genetic algorithm
Combine with BP algorithm and both can improve the accuracy of prediction it is also possible to improve convergence of algorithm speed.Genetic algorithm is first to god
It is optimized through network initial weight, orients preferable search space, then optimal value is searched in little space using BP algorithm.
The concretely comprising the following steps of genetic algorithm
1) initialize population
Randomly generate a population Xm×n, each individual X1×nRepresent the initial weight distribution of a neutral net, each gene
It is worth for a connection weight, then individual length is the number of neural network weight, that is,
N=r × s1+s1×s2+s1+s2
Wherein, n is individual length, and r is input layer number, s1For node in hidden layer, s2For output layer nodes
2) fitness function
According to fitness function value, individuality is evaluated, BP neural network input sample is obtained to each individual decoding,
Calculate output error value E, obtain fitness function f.
Calculate each individual adaptive value, population at individual fitness the maximum enters sub- population.
3) selection opertor
Using roulette method selection opertor.If the adaptive value f of i-th individualityi, then selected probability be
Wherein, m is population scale.
4) crossover operator
Crossover operator selects arithmetic crossover, goes out two new individualities by two individual linear combinations.Assume at two
Body Xi(k)、Xi+1With crossover probability p between (k)cCarry out crossover operation, then two new individuals producing after intersecting are
Wherein, Xi(k)、Xi+1K () represents i-th and the individual gene in kth position of i+1 respectively, α and β is between 0 and 1
Random number.
5) mutation operator
Select uniform mutation operator, to each genic value, with aberration rate pmCorresponding gene codomain takes a random number to enter
Row is replaced.
Xi=Xi(p)+r×q+Xi(n-p-1)
Wherein, q is the corresponding threshold width of+1 genic value of pth.
6) calculate fitness function
Calculate fitness function value, judge whether to reach maximum iteration time or meet required precision, otherwise return to step
2).
One .BP neural metwork training unit
BP neural network is a kind of multilayer feedforward neural network.The topological structure of 3 layers of BP network is as shown in figure 3, include defeated
Enter layer, hidden layer and output layer, each neuron is connected with next layer of all of neuron, with connectionless between layer neuron.Hidden
According to Kolmogorov theorem, if output layer has n nodes, the nodes of hidden layer are 2n+1 for setting containing layer.
In the present invention, the ultimate principle of BP neural network is to make network using gradient descent method adjustment weights and threshold value
The square mean error amount of real output value and desired output reduces.Standard BP algorithm revise weights when do not account for before the moment
Gradient direction so that learning process is it occur frequently that vibration, convergence is slow.Therefore the present invention adopts a kind of improved BP study
Algorithm, reduces the vibration trend of learning process, improves convergence by introducing momentum term.
The training process of BP neural network is as shown in figure 4, comprise the following steps that：
1) netinit.Determine network input layer nodes n, node in hidden layer l, output layer nodes m, input layer
Connection weight and hidden layer, hidden layer and output layer neuron between is wij, vjk, hidden layer threshold value a=[a1,a2,,al],
Output layer threshold value b=[b1,b2,,bm]
2) hidden layer output hj
Wherein, f is hidden layer excitation function, xiFor i-th input node variable.
3) output layer output ok
4) right value update
wij(t+1)=wij(t)+η[(1-β)D(t)+βD(t-1)]
vij(t+1)=vij(t)+η[(1-β)D'(t)+βD'(t-1)]
Wherein, η is learning rate, η ＞ 0,β is factor of momentum,
0≤β ＜ 1.
5) threshold value updates.
O is exported according to networkkWith desired output ykBetween error update aj, bk
bk(t+1)=bk(t)+(yk-ok)
6) whether iteration terminates evaluation algorithm, if being not over, return to step 2).
Two. Genetic BP Neutral Network predicting unit
As shown in figure 1, by front once arrive prediction during daily wind speed, precipitation, relative humidity, AQI summed data
ESDD data sub- with front primary insulation puts into Genetic BP Neutral Network predicting unit.In Genetic BP Neutral Network predicting unit,
The weights of the input layer that note BP neural network training unit obtains to hidden layer are Wij, the threshold value of hidden layer neuron is θj, hidden
Connection weight containing layer to output layer is Vij, the threshold value of output layer is
The then input of each neuron of hidden layer is
The transmission function of neutral net adopts Sigmoid function f (x)=1/ (1+e-x), then hidden layer is output as
Therefore obtain the input of output layer neuron and be output as
Three. prewarning unit
Prewarning unit arranges A, B, C, D, 4 advanced warning grades.Wherein when insulator ESDD numerical value reaches it may happen that pollution flashover
When insulator ESDD numerical value 95% when, i.e. 95% ρF, unit sends A level early warning；When insulator ESDD numerical value reaches and may send out
During raw pollution flashover insulator ESDD numerical value 90% when, i.e. 90% ρF, unit sends B level early warning；When insulator ESDD numerical value reaches
It may happen that during pollution flashover insulator ESDD numerical value 85% when, i.e. 85% ρF, unit sends C level early warning；When insulator ESDD number
Value reach it may happen that during pollution flashover insulator ESDD numerical value 80% when, i.e. 80% ρF, unit sends D level early warning.
The insulator being calculated by Genetic BP Neutral Network predicting unit ESDD prediction data is put into early warning criterion
Module, insulator ESDD prediction numerical value is contrasted by early warning criterion module with there is insulator ESDD numerical value during pollution flashover, to send pre-
Alarming information is processed for operations staff, thus timely and effectively preventing transmission line of electricity from the probability of pollution flashover accident occurring.
Claims (7)
1. a kind of equivalent salt density early warning system of electric insulator surface area dirt it is characterised in that：Including genetic algorithm training
Unit, BP neural network training unit, Genetic BP Neutral Network predicting unit, this four functional units of prewarning unit.
2. electric insulator surface area dirt as claimed in claim 1 equivalent salt density early warning system it is characterised in that：To lose
Propagation algorithm and BP algorithm combine and both can improve the accuracy of prediction, improve convergence of algorithm speed, genetic algorithm is first to god
It is optimized through network initial weight, orients search space, then optimal value is searched in little space using BP algorithm.
3. electric insulator surface area dirt as claimed in claim 2 equivalent salt density early warning system it is characterised in that：Heredity
BP neural network predicting unit by front once arrive prediction during daily wind speed, precipitation, relative humidity, AQI summed data
ESDD data sub- with front primary insulation puts into Genetic BP Neutral Network predicting unit.
4. electric insulator surface area dirt as claimed in claim 3 equivalent salt density early warning system it is characterised in that：Losing
Pass in BP neural network predicting unit, the weights of input layer to the hidden layer that note BP neural network training unit obtains are, hidden
Threshold value containing layer neuron is, the connection weight of hidden layer to output layer is, the threshold value of output layer is,
The then input of each neuron of hidden layer is
The transmission function of neutral net adopts Sigmoid function, then hidden layer be output as
Therefore obtain the input of output layer neuron and be output as
.
5. electric insulator surface area dirt as claimed in claim 4 equivalent salt density early warning system it is characterised in that：Early warning
Unit arranges A, B, C, D, 4 advanced warning grades, insulator ESDD numerical value wherein when insulator ESDD numerical value reaches generation pollution flashover
95% when, that is,, unit sends A level early warning；The insulator ESDD numerical value when insulator ESDD numerical value reaches generation pollution flashover
When 90%, that is,, unit sends B level early warning；The insulator ESDD numerical value when insulator ESDD numerical value reaches generation pollution flashover
When 85%, that is,, unit sends C level early warning；The 80% of insulator ESDD numerical value when insulator ESDD numerical value reaches generation pollution flashover
When, that is,, unit sends D level early warning.
6. electric insulator surface area dirt as claimed in claim 1 equivalent salt density early warning system it is characterised in that：BP god
It is a kind of multilayer feedforward neural network through network, the topological structure of 3 layers of BP network, including input layer, hidden layer and output layer, respectively
Neuron is connected with next layer of all of neuron, with connectionless between layer neuron, the setting of hidden layer according to
Kolmogorov theorem understands, if output layer has n nodes, the nodes of hidden layer are 2n+1.
7. a kind of equivalent salt density early warning system Forecasting Methodology of electric insulator surface area dirt as described in claim 1-6, its
It is characterised by：Comprise the following steps：
1）By sample data put into genetic algorithm functional unit, before sample data once arrive predict during daily wind speed, fall
The water yield, relative humidity, the summed data of AQI and a front ESDD data totally five parameters as input quantity, when secondary ESDD prediction
Data as output, through initialization population, the mistake that calculates fitness function, selection operation, crossover operation and mutation operation
Journey obtains the weights optimizing and threshold value；
2）The weights and threshold value that process the optimization obtaining from genetic algorithm training unit are put into BP neural network functional unit, makees
For initial weight and threshold value, in BP neural network functional unit, the equally former daily wind speed once arriving during prediction, precipitation
Amount, relative humidity, the summed data of AQI and a front ESDD data totally five parameters as input quantity, when secondary ESDD predicts number
According to as output, through the Error Calculation of hidden layer and output layer, weights and threshold value are constantly adjusted by training sample training,
Obtain best weight value and the threshold value of model；
3）The best weight value being obtained through sample training according to BP neural network functional unit and threshold value, set up insulator ESDD pre-
Survey model, by the front daily wind speed once arriving during prediction, precipitation, relative humidity, the summed data of AQI and a front ESDD
Numerical value puts into forecast model, output prediction insulator ESDD numerical value；
4）The insulator ESDD numerical value of prediction is put into early warning criterion unit, if ESDD numerical value reaches, and pollution flashover accident occurs
Threshold value then alert, the threshold value setting of pollution flashover flashover early warning is according to different gradation for surface pollution are regional, different electric pressures and
The sub- configuring condition of different insulative and determine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611005524.5A CN106405352A (en) | 2016-11-16 | 2016-11-16 | Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611005524.5A CN106405352A (en) | 2016-11-16 | 2016-11-16 | Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106405352A true CN106405352A (en) | 2017-02-15 |
Family
ID=59230446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611005524.5A Pending CN106405352A (en) | 2016-11-16 | 2016-11-16 | Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106405352A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108169243A (en) * | 2017-12-25 | 2018-06-15 | 国网宁夏电力有限公司固原供电公司 | Arrester pollution degree collecting unit scaling method based on BP genetic neural networks |
| CN108881660A (en) * | 2018-05-02 | 2018-11-23 | 北京大学 | A method of computed hologram is compressed using the quantum nerve network of optimization initial weight |
| CN109406380A (en) * | 2018-12-19 | 2019-03-01 | 国网北京市电力公司 | The detection method and device of insulator contamination accumulation characteristics under haze environment |
| CN109635919A (en) * | 2018-11-05 | 2019-04-16 | 新乡航空工业(集团)有限公司 | A kind of air class check valve valid circulation area measuring method and its device |
| CN111325325A (en) * | 2020-02-20 | 2020-06-23 | 贵州电网有限责任公司 | Method for predicting electric energy substitution potential based on genetic algorithm and BP neural network combination |
| CN112052939A (en) * | 2020-08-19 | 2020-12-08 | 国网山西省电力公司 | Active early warning system based on neural network algorithm |
| CN113095499A (en) * | 2021-03-26 | 2021-07-09 | 云南电网有限责任公司电力科学研究院 | Insulator equivalent salt deposit density prediction method |
| CN113252218A (en) * | 2021-05-12 | 2021-08-13 | 太原理工大学 | Insulator surface stress prediction method and prediction device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101893674A (en) * | 2010-07-12 | 2010-11-24 | 沈阳工业大学 | Pollution flashover index forecasting method for regional power grid |
| CN103411970A (en) * | 2013-07-17 | 2013-11-27 | 同济大学 | Alternating current transmission line insulator contamination condition detection method based on infrared thermography |
| CN103886395A (en) * | 2014-04-08 | 2014-06-25 | 河海大学 | Reservoir optimal operation method based on neural network model |
| US20160117845A1 (en) * | 2014-10-27 | 2016-04-28 | King Fahd University Petroleum and Minerals | Contamination level estimation method for high voltage insulators |
-
2016
- 2016-11-16 CN CN201611005524.5A patent/CN106405352A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101893674A (en) * | 2010-07-12 | 2010-11-24 | 沈阳工业大学 | Pollution flashover index forecasting method for regional power grid |
| CN103411970A (en) * | 2013-07-17 | 2013-11-27 | 同济大学 | Alternating current transmission line insulator contamination condition detection method based on infrared thermography |
| CN103886395A (en) * | 2014-04-08 | 2014-06-25 | 河海大学 | Reservoir optimal operation method based on neural network model |
| US20160117845A1 (en) * | 2014-10-27 | 2016-04-28 | King Fahd University Petroleum and Minerals | Contamination level estimation method for high voltage insulators |
Non-Patent Citations (1)
| Title |
|---|
| 张友鹏 等: "基于遗传BP神经网络的绝缘子泄漏电流预测", 《铁道学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108169243A (en) * | 2017-12-25 | 2018-06-15 | 国网宁夏电力有限公司固原供电公司 | Arrester pollution degree collecting unit scaling method based on BP genetic neural networks |
| CN108881660A (en) * | 2018-05-02 | 2018-11-23 | 北京大学 | A method of computed hologram is compressed using the quantum nerve network of optimization initial weight |
| CN109635919A (en) * | 2018-11-05 | 2019-04-16 | 新乡航空工业(集团)有限公司 | A kind of air class check valve valid circulation area measuring method and its device |
| CN109635919B (en) * | 2018-11-05 | 2023-01-31 | 新乡航空工业(集团)有限公司 | Method and device for measuring and calculating effective circulation area of air check valve |
| CN109406380A (en) * | 2018-12-19 | 2019-03-01 | 国网北京市电力公司 | The detection method and device of insulator contamination accumulation characteristics under haze environment |
| CN111325325A (en) * | 2020-02-20 | 2020-06-23 | 贵州电网有限责任公司 | Method for predicting electric energy substitution potential based on genetic algorithm and BP neural network combination |
| CN112052939A (en) * | 2020-08-19 | 2020-12-08 | 国网山西省电力公司 | Active early warning system based on neural network algorithm |
| CN113095499A (en) * | 2021-03-26 | 2021-07-09 | 云南电网有限责任公司电力科学研究院 | Insulator equivalent salt deposit density prediction method |
| CN113252218A (en) * | 2021-05-12 | 2021-08-13 | 太原理工大学 | Insulator surface stress prediction method and prediction device |
| CN113252218B (en) * | 2021-05-12 | 2023-11-17 | 国网山西省电力公司电力科学研究院 | Insulator surface stress prediction method and prediction device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106405352A (en) | Equivalent salt deposit density (ESDD) prediction and early warning system for power insulator surface contaminant | |
| Fazelpour et al. | Short-term wind speed forecasting using artificial neural networks for Tehran, Iran | |
| CN105117602B (en) | A kind of metering device running status method for early warning | |
| Bilal et al. | Wind turbine power output prediction model design based on artificial neural networks and climatic spatiotemporal data | |
| Minglei et al. | Classified real-time flood forecasting by coupling fuzzy clustering and neural network | |
| CN108280998A (en) | Short-time Traffic Flow Forecasting Methods based on historical data dynamic select | |
| CN112101669B (en) | Photovoltaic power interval prediction method based on improved extreme learning machine and quantile regression | |
| CN109919356B (en) | BP neural network-based interval water demand prediction method | |
| CN108764473A (en) | A kind of BP neural network water demands forecasting method based on correlation analysis | |
| Kolhe et al. | GA-ANN for short-term wind energy prediction | |
| CN115423301B (en) | Intelligent electric power energy management and control method, device and system based on Internet of things | |
| CN115935215B (en) | Power transmission line icing early warning method and system based on deep learning in extreme weather | |
| CN117498400B (en) | Distributed photovoltaic and energy storage data processing method and system | |
| CN106656357A (en) | System and method of evaluating state of power frequency communication channel | |
| Eseye et al. | Short-term wind power forecasting using a double-stage hierarchical hybrid GA-ANN approach | |
| CN109389238A (en) | A kind of short-term load forecasting method and device based on ridge regression | |
| CN106055579A (en) | Vehicle performance data cleaning system based on artificial neural network, and method thereof | |
| CN114118596A (en) | Photovoltaic power generation capacity prediction method and device | |
| CN106815635A (en) | A kind of forecasting system and method for insulator surface equivalent salt deposit density | |
| CN113988655A (en) | Power transmission line running state evaluation method considering multiple meteorological factors | |
| CN107632521A (en) | A kind of potentiostat control strategy based on decision tree and neutral net | |
| CN114357670A (en) | Power distribution network power consumption data abnormity early warning method based on BLS and self-encoder | |
| CN106503794A (en) | A kind of gear case of blower method for predicting residual useful life | |
| CN112434887B (en) | Water supply network risk prediction method combining network kernel density estimation and SVM | |
| CN115222160B (en) | Rail transit traction load prediction method based on measured big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170215 |
|
| RJ01 | Rejection of invention patent application after publication |