Minimum Distortion Point Tracking

Abstract

This article introduces the notion of minimum distortion point tracking (MDPT) : a control strategy where switching waveforms are optimally phase shifted to minimize aggregate ripple power for networks of dc–dc converters that are connected in series or parallel at the input or output. In a sense, MDPT generalizes the ubiquitous concept of interleaving in balanced systems to a broad class of asymmetric series- or parallel-connected dc–dc converters. For networks of up to one hundred interconnected power converters, MDPT demonstrates a one to two order of magnitude reduction (−14 to −22 dB) in distortion power. We present and experimentally verify three algorithms that can dynamically solve the MDPT optimization problem on a network of three input-parallel connected dc–dc buck converters handling 1.8 kW. The experimental results illustrate an up to $\text{3.06}\times$ reduction in the peak-to-peak ripple of the parallel-side bus voltage and convergence close to an optimal steady-state solution in 5 ms.