Modeling, Analysis, Simulation and Design Optimization (Genetic Algorithm) of dc-dc Converter for Uninterruptible Power Supply Applications

This paper presents the computer modeling, analysis, simulation and design optimization of an efficient dc-dc converter with synchronous rectification, which is important in battery charge/discharge circuits in dc uninterruptible power supply (UPS) systems. Simulation results from the models are presented for the forward mode and the backup mode. The results are compared to those of the expected ideal converter operation to confirm the validity of the simulation models. An objective function of the converter efficiency in the forward mode is derived and presented. Genetic Algorithm (GA) is used to perform a constrained optimization on the objective function and optimum results are presented. The resulting effects of GA parameters such as varying the population size and number of generations, and varying the amount of mutation and crossover fraction, are also presented. The computer modeling, analysis, simulation and design optimization (GA) are performed using software tools within the MATLAB/SIMULINK simulation environment.