Study On Harmonic Detection And Predictive Control For Shunt Active Power Filter

With the development and improvement of power electronic technology, a variety of power conversion devices are being used more widely, which seriously affected the quality of the modern power grid. Shunt active power filter has became an effective tool to solve this problem because of its better harmonic compensation performance.The subject aims to study the electromagnetic interference when the APF is operating and the control delay in order to improve the harmonics compensation performance of the APF.First, adaptive filters is used to detect the harmonic in the dissertation after analyzing and contrasting some popular methods.Because the precision of harmonic detecting is interfered seriously by the electromagnetic interference (EMI), a harmonic detection method using adaptive notch filters is proposed in this article.Instead of the output of adaptive filters by its estimation of the autocorrelation, the proposed algorithm can effectively sense the useful part from the output with uncorrelated measurement noise.At the same time, the algorithm can detect the harmonic current more accurately and make the adapt filter convergence more fast.Secondly, to solve the delay problem in the digital control system, a improved predictive current controller for active shunt power filters is proposed in this work.A discrete model of the system is establish,one step ahead prediction of reference voltage of the inveter is calculated by the model.To get the prediction value of the reference voltage, ASF current、reference current and the supply voltage need be predicted. The reference current and supply voltage are obtained using Lagrange interpolation formula, and ASF current is calculated by the reference current.After getting predictions of the reference voltage, the PWM singal is generated by means of the SVPWM technology.Finally, a three phase shunt active power filter was designed and built with a DSP TMS320F2812 processor from TI.The C program for the compensation algorithm is implemented and the parameters of the experimental system is improved.The experimental results show that the proposed algorithm makes the device have a good performance than before under the EMI and delay.