Decentralized PWM Interleaving for Ripple Minimization in both Symmetric and Asymmetric Parallel-connected DC-DC Converters

A decentralized control technique for minimizing the input and output current ripple in parallel connected dc-dc converters under both uniform and non-uniform operating conditions is presented in this paper. We solve two key problems related to obtaining minimum ripple operation of these converters. First, we eliminate the need for a centralized controller that manages system-wide pulse width modulation. This yields a decentralized structure where converters autonomously converge to the operating point with minimized ripple using only local voltage feedback. Second, we obtain a controller that can function in the presence of parametric mismatches as well as nonuniformities in converter duty ratios that naturally arise with heterogeneous input sources. More precisely, the proposed controller minimizes the fundamental switching harmonic under both uniform and asymmetric operating conditions. For the case studies we consider, a more than $4 \times$ reduction in net current ripple is observed compared to conventional methods. Moreover, the undesired fundamental harmonic is reduced by more than 30dB. Relevant analysis, simulation results and experimental results with 5 parallel-connected dc-dc converters validate the proposed method.