Neural network based AC-DC boost converter with improved power factor and regulated voltage

The switching control of a bidirectional AC-DC converter by Neural Networks is proposed in this paper. With this neural network application, an AC-DC converter has advantages such as quick switching response, simpler structure and better output waveform. The converter can be modified by a PWM rectifier switching state or predictive state observer and controller such as artificial intelligence based one to avoid time-delays causing transient conditions. The obtaining method of the circuit is given. The simulation of switching conditions is shown. To confirm the approximation and the operation by detecting and reducing line harmonic currents, the theoretical results and computer simulations are given.