Evaluation of Power Quality by Fuzzy Artificial Neural Network
Lin ZHOU,Qiu-hua LI,Hua-yong LIU,Feng ZHANG,Yong-jie QU
DOI: https://doi.org/10.3969/j.issn.1003-6520.2007.09.015
2007-01-01
Abstract:The synthetic quantified evaluation for power quality is one of the gist to weigh excellent or inferior of the power quality,and this is propitious to realize the electricity-purchase principle of discussing the price according to quality and high quality high price,thereby serve as to build the equitable electric power market.The problem of power quality evaluation is a desiderate problem.Also world countries have establish a series of the power quality standards,but this standards can only used to confirm the power quality is eligibility or not,can not confirm the power quality is excellent or not,therefore this standard witch only have two grades can not reflect the power quality entirely,reality and naturally.How to evaluate the power quality properly and based on this to deal with the problem of power quality reasonable and economically have been put on us imminently.Neural Network(NN)and Fuzzy Recognition(FR)applying in practice widely witch are used in the processing system to simulate the information of the organism.Each of both has its strong point,combining with each other,they can construct the Fuzzy Neural Recognition Network,and then the processing ability of the system could be enhanced.This paper first proposes to use the fuzzy neural network recognition which combining NN and FR for power quality evaluation,then build the Fuzzy Neural Network Recognition model(FNNR).This model not only calculates the power quality grade of the observation station,but also compares the differences between observation stations in the same grade.Fuzzy Neural Network Recognition simulates the thinking of the brain,has very strong self-organizing,self-learning,self-adapting ability.The model proposed in this paper has essential differences with the actual power quality integrate-evaluating methods;the output result of the model has the character of alone,objectivity,etc.Case results show that use the proposed model is objective and reasonable,with a special advantage when used in power quality evaluation.To evaluate the power quality properly to deal with the problem of power quality reasonably and economically,the fuzzy neural network recognition which combining NN and FR was put forward for power quality evaluation,then the Fuzzy Neural Network Recognition model(FNNR)was established.This model not only calculates the power quality grade of the observation station,but also compares the differences between observation stations in the same grade.Fuzzy Neural Network Recognition simulates the thinking of the brain,and has very strong self-organizing,self-learning,self-adapting ability.The model proposed in this paper has essential differences with the actual power quality integrate-evaluating methods;the output result of the model has the characteristic of sole result and objectivity,etc.Case results show that the proposed model is objective and reasonable with a special advantage used in power quality evaluation.