Research Of Unified Power Quality Conditioner Based On Neural Network

With the continuous development of power electronics technology, the use of active power filter to achieve harmonic suppression and reactive power compensation has gradually become one of the active research areas of power electronics. This paper mainly studies the Unified Power Quality Conditioner based on neural network, which can effectively solve multiple power quality problems either form the load or form the line side. This paper firstly concentrates on topological structure and the working principle of UPQC. Then, the detection method and the control strategy, which are the important parts affecting the compensation effect of the UPQC, are discussed and investigated in detail in the thesis. Discussion and analysis of traditional instantaneous reactive power theory advantages and disadvantages of harmonic detection methods, in this based on the theory of neural network based detection methods, this method is more simple and can improve the detect accuracy. In additional, Simulation results given in the thesis verify the correctness of the harmonic detection method and the control strategy proposed in the thesis.