PD-type iterative learning control with forgetting factor for phase-shift full-bridge series resonant inverters

Abstract

This paper addresses the fixed initial offset problem of iterative learning control (ILC) for phase-shift full-bridge series-resonant inverters (PSFB-SRIs). Considering the specific characteristics of PSFB-SRIs, a novel PD-type learning algorithm with forgetting factor is proposed. It is shown that the algorithm guarantees the output tracking errors are bounded in the presence of state disturbances, and asymptotic convergence to zero in their absence. Furthermore, an initial rectifying mechanism is introduced to eliminate the effect of fixed initial offsets. When the learning algorithm with the initial rectifying term is applied, the output tracking errors can tend to zero within a pre-specified interval under disturbance-free conditions. In the end, a simulation example is given to demonstrate the effectiveness of the proposed method.