I. Introduction
The dc-link capacitor bank occupies a large volume in power electronic systems of the electric vehicles; sometimes it is more than 40% of the traction inverter [1]. Hence, the volume of the dc-link capacitor greatly affects the system power density. Most of the existing work on dc-link capacitance minimization follows either one of two major approaches, i.e., in the first one, the capacitor minimization is realized using the system controller [2]-[5]; in the second one, the required capacitance is minimized by matching the switching algorithm of the dc-dc converter and the inverter; and the interaction between them [6]–[7]. In all cases, a highly-robust control system becomes extremely important as the capacitor value is decreased towards its stability limit. Closed-loop control methodology for a three-phase back-to-back ac-dc-ac converter was proposed in [5], in which the dc-link capacitor reduction was realized by balancing the power generation and consumption. An optimization method to determine the minimum current through the dc-link capacitor when the switching frequency of the dc-dc converter and the inverter are equal is explored in [6]. But, using a look-up table in every cycle makes the control system complicated. Furthermore, higher switching frequencies for the dc-dc converter could shrink the passives leading to higher power density. Most of these references, however, do not provide a formal methodology for calculating the minimum dc-link capacitor value.