ANN Control for Switched Quasi Sepic Converter Integrated VSI for a Grid Connected System

In this research, Artificial Neural Network (ANN) based control is implemented to control the Switched Quasi Sepic Converter (SQSC) for a $1 \varphi$ grid connected system. The design of a $1 \varphi$ voltage source inverter (VSI) using proportional integral control for pulse width modulation (PWM) is used to eliminate harmonics generated during the turn on and off of the inverter switches and compared with ANN at last. ANN control technique is implemented here in this research work for the SQSC, to highly improve the converter output voltage and to minimize the power quality issues. The intelligent control for SQSC depreciates the switching losses, and hence improves the power factor of the SQSC integrated grid connected system. Here, output voltage of SQSC converter output is enhanced with ANN control, by minimizing the switching losses in the controlled switches. LC filter smoothens the $1 \varphi$ inverter output and decreases the ripple content in the $1 \varphi$ inverter voltage output. The performance of this bidirectional converter is investigated in MATLAB/Simulink software and a hardware developed using the DSPIC30F4011 Microcontroller. The above setup finds application in electric vehicles, pv power fed induction motors for agricultural applications.