CN105205532A - Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure - Google Patents
Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure Download PDFInfo
- Publication number
- CN105205532A CN105205532A CN201510537557.3A CN201510537557A CN105205532A CN 105205532 A CN105205532 A CN 105205532A CN 201510537557 A CN201510537557 A CN 201510537557A CN 105205532 A CN105205532 A CN 105205532A
- Authority
- CN
- China
- Prior art keywords
- insulator
- neural network
- grading ring
- ring
- optimization method
- 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
Landscapes
- Insulators (AREA)
Abstract
The invention relates to a neural-network-genetic-algorithm-based optimization method of an insulator with a grading ring structure. A neural network model is established by using a grading ring structure and position as input variables and an insulator along surface field intensity and a maximum field intensity of a grading ring surface as objective functions; a mapping relation between the input variables and the objective functions is established and the mapping relation is expressed as follows: F(E1, E2)=F(R, r, h), wherein the E1 expresses the maximum field intensity of the along surface of the insulator, the E2 expresses the maximum field intensity of the surface of the grading ring, and the R, r, and H express the ring radius, the pipe radius, and the covering depth of the grading ring. According to the invention, a neural-network-genetic-algorithm-based optimization model is established and the grading ring structure and position are used as input variables and the insulator along surface field intensity and the maximum field intensity of the grading ring surface are used as objective functions, so that the insulator structure is optimized. Therefore, an objective of line safety keeping can be achieved.
Description
Technical field
The invention belongs to technical field of electric system protection, relate to a kind of insulator structure optimization method, be specifically related to a kind of insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm.
Background technology
Along with development that is super, extra-high voltage large capacity transmission technology, composite insulator is applied to transmission line of electricity more and more.Composite insulator than traditional porcelain and glass insulator have anti-soil dodge ability strong, exempt from cleaning, the advantage such as lightweight, easy to maintenance, particularly there is the area of heavy pollution at China's transmission line of electricity, preferentially will adopt composite insulator.But, composite insulator due to Electric Field Distribution extremely uneven, often subject too high electric field near plug high-pressure side, the major accident such as easily cause core brittle fracture, go offline.Therefore, study the potential and electric field distributions of composite insulator and find rational Optimized Measures the safe operation of transmission line of electricity and maintenance are had great importance.
Summary of the invention
In order to overcome the deficiencies in the prior art, the invention provides a kind of insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm, by setting up the Optimized model based on neural network and genetic algorithm, using the structural parameters of grading ring and position as input variable, with insulator along face maximum field strength and grading ring surface maximum field strength for objective function, optimize insulator structure, reach the object of maintenance line safety.
Technical scheme of the present invention is: a kind of insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm, set up with Equalized voltage ring construction and position as input variable, with the neural network model that insulator is objective function along face field intensity and grading ring surface maximum field strength, set up the mapping relations between input variable and objective function, i.e. F (E
1, E
2)=F (R, r, h), in formula, E
1for insulator is along face maximum field strength, E
2for grading ring surface maximum field strength, R, r, h be respectively grading ring ring radius, pipe radius, cover into the degree of depth.Described grading ring surface maximum field strength E
2maximal value E
2maxmeet bloom field intensity, i.e. E
2max=g (R, r, h)≤2.2kV/mm.Described train samples is obtained by Finite element arithmetic, then the output valve of network is regarded as individual adaptive value, the best input value of searching system is carried out with the genetic operator such as selection, intersection, variation of genetic algorithm, make corresponding optimal adaptation angle value optimum, thus find best Equalized voltage ring construction parameter.The field intensity value of described grading ring parameter and output is interval in [0,1] normalization, and method for normalizing is: set f as grading ring arbitrary structures parameter, f
maxand f
minthe bound of parameter for this reason, for arbitrary f (x), if these structural parameters and E
1positive correlation, then the structural parameters f after normalization
n(x) be:
if these structural parameters and E
1negative correlation, then the structural parameters f after normalization
n(x) be:
described insulator is along face maximum field strength E
1constant interval span be [0.12.5] kV/mm, grading ring surface maximum field strength E
2interval scope be [0.13.1] kV/mm.The activation function of the hidden layer of described neural network selects Sigmoid function:
wherein y is exponential, determines the steepness of Sigmoid function.The training algorithm of described neural network selects LM algorithm.Described insulator high-pressure side is provided with two grading rings, and little endless tube radius is 50mm, ring radius is 400mm, raise apart from being 250mm; Large endless tube radius 120mm, ring radius being 1120mm, raising apart from being 500mm.Described insulator low pressure end is provided with grading ring, and low pressure end grading ring pipe radius is 120mm, and ring radius is 720mm, raises apart from being 9480mm.
The present invention has following good effect: by setting up the Optimized model based on neural network and genetic algorithm, using the structural parameters of grading ring and position as input variable, with insulator along face maximum field strength and grading ring surface maximum field strength for objective function, optimize insulator structure, reach the object of maintenance line safety.
Accompanying drawing explanation
Fig. 1 is the grading ring optimization neural network model of the specific embodiment of the invention.
Embodiment
Contrast accompanying drawing below, by the description to embodiment, the specific embodiment of the present invention is as the effect of the mutual alignment between the shape of involved each component, structure, each several part and annexation, each several part and principle of work, manufacturing process and operation using method etc., be described in further detail, have more complete, accurate and deep understanding to help those skilled in the art to inventive concept of the present invention, technical scheme.
Neural network model by learning and train input, exporting weights and structure that data just can obtain network, thus draws the relation lain in input, output data.This relation lies in neural network inside, does not need to know concrete model, only just need can approach multidimensional nonlinear characteristic between input and output by neural network, thus set up the funtcional relationship between constrained input.
Usually when being optimized design to grading ring, with insulator along face maximum field strength and grading ring surface maximum field strength for objective function, with Equalized voltage ring construction and position for input variable, be a kind of mapping relations between them, i.e. F (E
1, E
2)=F (R, r, h), in formula, E
1for insulator is along face maximum field strength, E
2for grading ring surface maximum field strength, R, r, h be respectively grading ring ring radius, pipe radius, cover into the degree of depth.Grading ring surface maximum field strength E
2maximal value E
2maxmeet bloom field intensity, i.e. E
2max=g (R, r, h)≤2.2kV/mm.Insulator is made to reach minimum along face maximum field strength, i.e. min (E
1max)=min (F (R, r, h)).
The present invention first calculates some samples by Finite Element Method, utilizes neural network R, r, h and E
1, E
2between mapping relations, then the output valve of this network is regarded as individual fitness value, the best input value of searching system is carried out with the genetic operator such as selection, intersection, variation of genetic algorithm, make corresponding fitness value optimum, thus finding optimum Equalized voltage ring construction parameter, the method is that the optimization problem of insulator high and low pressure side grading ring parameter provides new approaches.
The input and output of artificial neural network should choose nondimensional vector, so the field intensity value of the grading ring parameter of input and output is normalized in [0,1] interval, normalized is carried out as follows.Method for normalizing is:
If f is grading ring arbitrary structures parameter, f
maxand f
minthe bound of parameter for this reason, for arbitrary f (x), if these structural parameters and E
1positive correlation, then the structural parameters f after normalization
n(x) be:
If these structural parameters and E
1negative correlation, then the structural parameters f after normalization
n(x) be:
For the normalized of output field intensity values, need by Finite Element Method obtain for the E corresponding to the sample of training
1, E
2determine, insulator is along face maximum field strength E
1constant interval span be [0.12.5] kV/mm, grading ring surface maximum field strength E
2interval scope be [0.13.1] kV/mm.The activation function of the hidden layer of described neural network selects Sigmoid function:
wherein y is exponential, determines the steepness of Sigmoid function.The number of hidden layer neuron is by (2N+1) rule interestingness (N is the nodes of input layer), and neuron number is 7 herein.The training algorithm of neural network selects LM algorithm.
Concrete, insulator high-pressure side of the present invention is provided with two grading rings, and little endless tube radius is 50mm, ring radius is 400mm, raise apart from being 250mm; Large endless tube radius 120mm, ring radius being 1120mm, raising apart from being 500mm.Insulator low pressure end grading ring pipe radius is 120mm, and ring radius is 720mm, raises apart from being 9480mm.
Above by reference to the accompanying drawings to invention has been exemplary description; obvious specific implementation of the present invention is not subject to the restrictions described above; as long as have employed the improvement of the various unsubstantialities that method of the present invention is conceived and technical scheme is carried out; or design of the present invention and technical scheme directly applied to other occasion, all within protection scope of the present invention without to improve.
Claims (9)
1. the insulator optimization method based on the band Equalized voltage ring construction of neural network and genetic algorithm, it is characterized in that, set up with Equalized voltage ring construction and position as input variable, with the neural network model that insulator is objective function along face field intensity and grading ring surface maximum field strength, set up the mapping relations between input variable and objective function, i.e. F (E
1, E
2)=F (R, r, h), in formula, E
1for insulator is along face maximum field strength, E
2for grading ring surface maximum field strength, R, r, h be respectively grading ring ring radius, pipe radius, cover into the degree of depth.
2. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 1, is characterized in that, described grading ring surface maximum field strength E
2maximal value E
2maxmeet bloom field intensity, i.e. E
2max=g (R, r, h)≤2.2kV/mm.
3. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 2, it is characterized in that, described train samples is obtained by Finite element arithmetic, then the output valve of network is regarded as individual adaptive value, carry out the best input value of searching system with the genetic operator such as selection, intersection, variation of genetic algorithm, make corresponding optimal adaptation angle value optimum, thus find best Equalized voltage ring construction parameter.
4. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 3, is characterized in that, the field intensity value of described grading ring parameter and output is interval in [0,1] normalization, and method for normalizing is:
If f is grading ring arbitrary structures parameter, f
maxand f
minthe bound of parameter for this reason, for arbitrary f (x), if these structural parameters and E
1positive correlation, then the structural parameters f after normalization
n(x) be:
If these structural parameters and E
1negative correlation, then the structural parameters f after normalization
n(x) be:
5. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 4, is characterized in that, described insulator is along face maximum field strength E
1constant interval span be [0.12.5] kV/mm, grading ring surface maximum field strength E
2interval scope be [0.13.1] kV/mm.
6. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 4, is characterized in that, the activation function of the hidden layer of described neural network selects Sigmoid function:
wherein y is exponential, determines the steepness of Sigmoid function.
7. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 6, is characterized in that, the training algorithm of described neural network selects LM algorithm.
8. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 1, it is characterized in that, described insulator high-pressure side is provided with two grading rings, and little endless tube radius is 50mm, ring radius is 400mm, raise apart from being 250mm; Large endless tube radius 120mm, ring radius being 1120mm, raising apart from being 500mm.
9. the insulator optimization method of the band Equalized voltage ring construction based on neural network and genetic algorithm according to claim 8, it is characterized in that, described insulator low pressure end is provided with grading ring, and low pressure end grading ring pipe radius is 120mm, ring radius is 720mm, raises apart from being 9480mm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510537557.3A CN105205532A (en) | 2015-08-26 | 2015-08-26 | Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510537557.3A CN105205532A (en) | 2015-08-26 | 2015-08-26 | Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105205532A true CN105205532A (en) | 2015-12-30 |
Family
ID=54953201
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510537557.3A Pending CN105205532A (en) | 2015-08-26 | 2015-08-26 | Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105205532A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205900A (en) * | 2016-08-24 | 2016-12-07 | 国网江西省电力公司南昌供电分公司 | A kind of gold utensil method for designing improving insulator contamination voltage and gold utensil |
| CN110737998A (en) * | 2019-09-25 | 2020-01-31 | 中国电力科学研究院有限公司 | equalizing ring optimization design method based on finite element and depth belief network |
| CN111094956A (en) * | 2017-09-22 | 2020-05-01 | 沙特阿拉伯石油公司 | Processing the thermographic image with a neural network to identify Corrosion Under Insulation (CUI) |
| CN117272756A (en) * | 2023-11-14 | 2023-12-22 | 江苏勇龙电气有限公司 | Design method of equalizing ring for GIL combined type ultrahigh-voltage casing pipe |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2212826Y (en) * | 1995-03-07 | 1995-11-15 | 沈阳市东方铝材制品厂 | Synthesized isolator aluminium equalizing ring |
| CN2232157Y (en) * | 1995-05-19 | 1996-07-31 | 襄樊电力设备厂 | Equalizing rings |
| EP2605358A1 (en) * | 2010-06-30 | 2013-06-19 | State Grid Corporation of China | Wire connecting method and converter station for ultra-high voltage direct current power transmission, and ultra-high voltage direct current power transmission system |
-
2015
- 2015-08-26 CN CN201510537557.3A patent/CN105205532A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2212826Y (en) * | 1995-03-07 | 1995-11-15 | 沈阳市东方铝材制品厂 | Synthesized isolator aluminium equalizing ring |
| CN2232157Y (en) * | 1995-05-19 | 1996-07-31 | 襄樊电力设备厂 | Equalizing rings |
| EP2605358A1 (en) * | 2010-06-30 | 2013-06-19 | State Grid Corporation of China | Wire connecting method and converter station for ultra-high voltage direct current power transmission, and ultra-high voltage direct current power transmission system |
Non-Patent Citations (1)
| Title |
|---|
| 司马文霞等: "特高压复合绝缘子电场计算及基于神经网络遗传算法的均压环结构优化设计", 《高电压技术》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205900A (en) * | 2016-08-24 | 2016-12-07 | 国网江西省电力公司南昌供电分公司 | A kind of gold utensil method for designing improving insulator contamination voltage and gold utensil |
| CN111094956A (en) * | 2017-09-22 | 2020-05-01 | 沙特阿拉伯石油公司 | Processing the thermographic image with a neural network to identify Corrosion Under Insulation (CUI) |
| CN110737998A (en) * | 2019-09-25 | 2020-01-31 | 中国电力科学研究院有限公司 | equalizing ring optimization design method based on finite element and depth belief network |
| CN110737998B (en) * | 2019-09-25 | 2022-07-26 | 中国电力科学研究院有限公司 | Grading ring optimization design method based on finite element and deep belief network |
| CN117272756A (en) * | 2023-11-14 | 2023-12-22 | 江苏勇龙电气有限公司 | Design method of equalizing ring for GIL combined type ultrahigh-voltage casing pipe |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105205532A (en) | Neural-network-genetic-algorithm-based optimization method of insulator with grading ring structure | |
| Wang et al. | Flood control operations based on the theory of variable fuzzy sets | |
| CN104915515A (en) | BP neural network based GFET modeling method | |
| CN107168292B (en) | Submarine navigation device circuit failure diagnosis method based on ELM algorithm | |
| CN107967571A (en) | Index system generation method based on entropy assessment | |
| CN110245390A (en) | A kind of automobile engine oil consumption prediction technique based on RS-BP neural network | |
| CN108416178A (en) | A kind of mechanical elastic vehicle wheel Design of Structural parameters method | |
| CN104200038B (en) | The Preisach model optimum design method of electrical sheet magnetic hystersis loss | |
| CN108562495A (en) | A kind of computational methods of intact steel pipe limit internal pressure | |
| GA et al. | Existence and non-existence results for nonlocal elliptic systems via sub-supersolution method | |
| CN117272756A (en) | Design method of equalizing ring for GIL combined type ultrahigh-voltage casing pipe | |
| Bu et al. | An improved genetic algorithm for searching for pollution sources | |
| CN105512383A (en) | Dredging process regulation and control parameter screening method based on BP neural network | |
| CN110109904B (en) | Environment-friendly big data oriented water quality soft measurement method | |
| Yan et al. | An evaluation model of water quality based on DSA-ELM method | |
| CN106228241A (en) | The ore deposit rock intensity flexible measurement method that adaptability artificial bee colony optimizes | |
| CN104346501A (en) | Integrated optimization method and system for static model of fully-extending boom of crane | |
| Liu et al. | Existence of unbounded positive solutions for BVPs of singular fractional differential equations | |
| Zhou et al. | Soft soil foundation settlement prediction and economic cost management analysis based on new algorithm | |
| CN105046355A (en) | System energy consumption modeling method based on improved mind evolutionary algorithm | |
| Gao et al. | Summary study on the development of water assessment | |
| CN115936309B (en) | Building health monitoring method and system based on multi-feature factor integrated evaluation | |
| Xiao et al. | Dam Deformation Prediction Model Based on Combined Gaussian Process | |
| 郭锐 | Research on adaptive anti-rollover control of kiloton bridge transporting and laying vehicles | |
| Xiao et al. | Systematic Combat Effectiveness Evaluation Model Based on Xgboost |
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: 20151230 |
|
| RJ01 | Rejection of invention patent application after publication |