A new single-loop method for steady-state analysis and design of networks with switching power converters

A common strategy for steady-state analysis of switching power converters has been to pose the analysis problem as a nonlinear boundary-value problem which can be solved using time-domain shooting methods. However, the resulting nonlinear equations for closed loop converters are, in general, ill-conditioned. The proposed method uses a new set of constraints which define the nonlinear boundary-value problem. This new mapping results in substantial convergence improvement and, in addition, it allows for computation of converter steady-state in a single Newton's iteration loop. The problems of convergence failures due to the initial guess referred in the past are explained. The proposed method has been verified using several hard and soft-switching closed-loop converters.