Research On Fast Calculation Of Voltage Stability Margin Of The System Contained Wind Farm

Rapid development of the power system of generation,transmission and transformation electricity and the applications of a variety of new technologies make the system structure more and more complex.And constraints of transmission equipment and environmental and other factors,the transmission power of the power system closer and closer to the limit,the static voltage stability problem has become a hot issue.A lack of energy and environmental pollution problems prompted our country must vigorously develop renewable energy sources.Wind power is the main force of renewable energy,our country has developed greatly.But power system of my country become more weak after injecting the wind farm.In the power system,voltage collapse accidents have the features of latent and sudden,accidents are irreparable once happened.Therefore,in order to timely warning,so as to prevent accidents,on-line monitoring system voltage stability degree is imperative.How to make online judgment on the system voltage stability degree,calculate system voltage stability margin index quickly and accurately,get enough time to take measures for operating crew.This become focus in the study of static voltage stabilit y problem.Aim at fast calculation of voltage stability margin index,this paper has carried out the following studies:First of all,the paper analyzes the principle of continuation power flow method to calculate margin index.And uses the random disturbance model to simulate the actual operation state of the system,and calculated system margin index of various kinds of random running state.The margin index from continuation power flow method acts as samples of the neural network in the follow-up work.Secondly,this paper studies the problem of wind farm switching in power system,analyzes the calculation model of continuous power flow for wind indu ction generator,and the doubly-fed wind power generator regards as PQ node.To provide voltage stability margin index samples under all conditions of wind speed for neural network,this paper presents a random disturbance model of wind speed.Thirdly,generalized regression neural network(GRNN)is compared with the BP neural network(BPNN),then this paper puts forward using GRNN to calculate voltage stability margin index quickly.Due to high dimension of power system and the strong correlation of input samples,this paper uses principal component analysis(PCA)for dimension reduction of input samples,this paper proposes a PCA-GRNN model to compute margin index.In practice,in order to provide input samples of neural network,PMUs are used to measure the voltage vector on-line.Optimal PMUs placement is putted forward to reduce the number of PMUs,voltage vectors are acted as input samples of the neural network after the system becoming observable.In order to further improve efficiency,reduce the number of installed PMUs,and ensure the calculation precision,the article puts forward a method of not completely observable optimal placement based on sensitivity of margin index.Part of the node voltage vectors are chosen as input by graham schmidt ortho gonalization and neural network sensitivity technology,and the method of optimal PMUs are used to make the selected nodes observable.Finally,this paper takes IEEE30 in matpower software system for simulation which conclude normal systems,N-1 system,and system contained wind farms.The simulation results demonstrate the advantages of the presented method,and its reliability in practical application.