Research On Power System State Estimation Based On Adaptive Extended Set-Membership Filter Under Hybrid Measurement

Power system state estimation is the basis of power flow calculation,fault diagnosis,power quality analysis and operation scheduling.Power system state estimation can eliminate the noise interference in the measurement data by filtering the obtained measurement data,so as to provide accurate data support for the safe and stable operation of power system.It is an indispensable core part of energy management system(EMS).Considering that the current measurement data of power system is mainly provided by Supervisory Control and Data Acquisition(SCADA)and Wide Area Measurement System(WAMS),and the traditional state estimation method is not satisfied when the noise distribution is non-Gaussian.Therefore,this thesis will mainly study the application of adaptive extended set-membership filter(AESMF)in power system state estimation under SCADA/WAMS hybrid measurement.First of all,this thesis makes equivalent simplification of power equipment such as transmission lines,transformers,power networks,etc.,gives equivalent calculation model of power system,and establishes the mathematical model of dynamic state estimation of power system.At the same time,the weighted least square(WLS)algorithm and extended Kalman filter(EKF)algorithm,the two most commonly used filtering algorithms in power system state estimation,are briefly introduced,and the basic principle of set-membership filtering(SMF)algorithm are emphatically described.Then the advantages and disadvantages of these algorithms are summarized and compared.Then,aiming at the difference of SCADA/WAMS hybrid measurement data,two compatibility processing methods are proposed.For the problem of data composition difference,a two-stage mixed model is proposed,the first stage is to estimate SCADA measurement data,the second stage is to estimate the fused WAMS data and the first stage estimation results.For the problem of inconsistent data update frequency,the interpolation method and curve fitting ideas in numerical analysis are used to first interpolate SCADA short-term historical data to obtain a smaller time interval,and then use curve fitting to fit these discrete data,filling in the gap of WAMS data.Secondly,this thesis presents an integer programming optimal PMU placement method considering N-1 criterion,which has better data redundancy under the condition of satisfying power system observability.Finally,considering the poor adaptability of conventional state estimation algorithm to non-Gaussian noise and the practical application of set-membership filtering algorithm in power system state estimation,combined with the theory of adaptive algorithm,an adaptive extended set-membership filtering algorithm is proposed.By introducing the adaptive program,the measurement data can be preprocessed adaptively,which can reduce the possibility of iterative divergence and improve the estimation accuracy of the filtering algorithm.The simulation results show that the adaptive extended set-membership filtering algorithm can reduce the estimation error,shorten the estimation time,and has excellent estimation performance.