Genetic optimisation of interval mathematics-based sensitivity analysis of switching converters

Interval mathematics (IM) has been revealed in the recent past as a promising tool for the sensitivity analysis of circuits with uncertain parameters. In this paper, a genetic approach is proposed to improve the efficiency of IM-based sensitivity analysis of switching power converters. A genetic algorithm is used to establish an optimal uncertainty intervals partitioning to obtain a sharp evaluation of the sensitivity of any objective function with respect to uncertain parameters. It is shown that the joint IM-GA approach ensures the possibility of performing efficient and reliable small-change as well as large-change sensitivity analysis. Some applications to the analysis of switching power converters are proposed