The Application of a Novel Inverter with Fuzzy Neural Network Control for Rectifier Loads

• Main Authors: Cheng-Ho Yang, 楊政和
• Other Authors: Nan-Ming Chen
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/64835899673554503608

碩士 === 國立臺灣科技大學 === 電機工程系 === 102 === If the internal inverter encounters rectified loads when an AC power supply is used, it will draw a large instantaneous current and will make filter storage energy inadequate, leading to a distorted voltage waveform so that the total harmonic distortion will be increased and the quality of the power supply is affected. To deliver a fast response to load in the converter control, it often requires complex control algorithms. A digital signal processor not only can adopt the control law, but also has the advantage of avoiding the effect of component degradation as an analog controller. This study uses a digital signal processor (DSP TMS320F28335). By adopting software Psim and combining the control law of neural fuzzy control, based on past error information, it can appropriately correct the sinusoidal waveforms and thereby solve the problem of inadequate energy storage for the filter. The inverter output voltage waveform, therefore, can be approximate to a sinusoidal waveform. Experiments results show that this method can reduce the total harmonic distortion to around 3%, which can effectively improve the efficiency and quality of the power supply.