Estimation Of Dynamic Harmonics In Power Systems Based On Particle Swarm Optimized Recursive Least Square Model

With the development of electric power system,nonlinear loads have been increased in power system.harmonic problems are becoming more and more prominent.Scholars have made a fairly thorough study of it and the good result is obtained.There are various of methods used in the power system in the world,like Fourier transform(FT),fast Fourier transform(FFT),artificial neural network,wavelet transform,Kalman filter(KF)and least squares algorithm and so on,and their test performance and applicability are discussed in this paper.First,power quality analysis methods are introduced,such as root mean square,Fourier transform,wavelet transform,Kalman filter and the recursive least square,its simulation analysis,merit and demerit has been given.Then,This paper presents a method for estimating harmonics in power systems based on particle swarm optimized recursive least square(PSO-RLS)model,PSO is used to get the optimum initial weights,and RLS is used to updates the weights of the harmonic signal,the method resolves the problem that RLS is sensitive to initial weights and performance of RLS in harmonic estimation was optimized.The stationary voltage signal and dynamic voltage signal are simulated using the proposed algorithm,and its performance is compared in difference noise environment.Finally,the dynamic sub harmonic and inter harmonics are simulated.Simulation results show that the proposed algorithm have a better performance compared with Variable Constraint based Least Mean Square and GA-RLS algorithm.