Genetic Algorithm Based Tuning of Nonfragile and Robust PI Controller for PSFB DC-DC Converter

摘要

Phase shift full bridge (PSFB) DC-DC converters are used to convert high voltage to low voltage for medium and high power applications. In this paper a PSFB DC-DC converter is designed for converting 350 V DC supply to a constant output voltage of 12 V with a maximum load of 2 kW. It is desired to satisfy a stringent set of time domain specifications on settling time, steady state error and peak overshoot (to be less than 50 ms, 200 mV and 25% respectively) over a wide range of operating conditions. To obtain this objective, we propose a genetic algorithm (GA) based PI controller tuning for PSFB DC-DC converter. Simulation results show that $K_{p}$ and $K_{i}$ values that obtained using GA help to vary the load from 100 W to 2000 W, with steady state error, settling time and peak overshoot less than the specified limit. The system also supports an input voltage range from 300 V to 500 V and an output voltage from 8 V to 18 V. Fragility of the controller is analysed and the results show that controller is nonfragile.