Research On Application Of CS Algorithm In Harmonic Detection And Estimation

Cuckoo search algorithm(CS) is one of promising and widely used optimization algorithm. Now it has been commonly used in multi-objective optimization, identification of parameter, computer network and so on. But for power system, CS algorithm has rarely been used. With a large number of nonlinear electronic devices and other components increasing in the circuit, the harmonics generated in the power system voltage and current waveforms are easily distort the wave form. Therefore, as a harmonic suppression of main parts, accuracy and real time of harmonic detection become particularly important. So the improved CS algorithm which has higher accuracy is applied to harmonic estimation. Secondly,parameters of Kalman filter(KF) were optimized by the improved cuckoo search algorithm to achieve goals of high precis ion, better real time.Firstly, based on the analysis of the working principle of the CS algorithm, an improved method combine with the chaos theory is proposed to the artic le. It is mainly applied to chaos initialization, and then chaotic disturbance is added to the local optimal to improve the accuracy and convergence speed of the algorithm. By achieving the programming of 5 test functions, the improved CS algorithm is proved to have higher convergence precision and faster convergence speed.Secondly, researching harmonic estimation model determines the mean square error between actual value and estimated value as fitness function of the cuckoo search algorithm for harmonic estimation. Compared to experimental data of partic le swarm optimization algorithm, improved cuckoo search algorithm has better performance in simulation experiments and results of harmonic estimation.In addition, working princ iple of the traditional Kalman Filter algorithm is presented in my proposal. Values of System noise and measurement noise covariance directly affect the performance of the algorithm, so cuckoo search optimized kalman filter is proposed to determine squared error between estimated value and the actual value as fitness function. Then the system noise and the measurement noise covariance, which are the two main parameters are optimized. The results of the simulation experiments show that the optimized KF detection algorithm has higher detective precis ion and the real time performance than before.At last, improved CSKF detection method is applied to Active Power Filter(APF). And compared with compensation effect of harmonic detection based on traditional KF, the improved one is proved to directly influence compensation effect of APF and effectively improve it for optimization of the Kalman Filter harmonic detection method.