A multi-objective optimization of a Resonant Boost - Half - Bridge Converter aimed at solar residential air conditioning considering site climatic factors

Torrid temperatures in northwestern México have caused the widespread installation of air-conditioning appliances which, in turn, lead to a high demand of electric energy. Peak consumption, however, occurs when the irradiance is at its highest, suggesting that the problem can be partially solved by injecting photovoltaic energy into the DC link of the adjustable speed drive in inverter-grade air-conditioning appliances. In this paper, the optimal design of an isolated Resonant Boost - Half - Bridge DC/DC Converter is described. The converter was optimized using the NSGA-II algorithm, with weighted-efficiency and cost as objective functions. Annual irradiance and temperature data was used to develop site-specific weighted-efficiency coefficients. Since photovoltaic power evolves throughout the day, several optimizations, at different power levels, were carried out. Components to build the converter were selected from a data base taking into account individual temperatures, and suitable maximum and peak ratings. Optimization results demonstrate that optimizing the converter at its nominal power does not provide the best solution. Rather, it was found that the best result is obtained when the converter is designed at the power level that coincides with the maximum energy harvesting. The main differences between optimization results are in the transformer design, leakage inductance, switching frequency and capacitors. Results also highlighted the fact that cost is very sensitive to efficiency.