Three-Level Optimized Pulse Patterns for Grid-Connected Converters with LCL Filters

This paper presents the computation of three-level optimized pulse patterns (OPPs) for converters connected to the grid via LCL filters. To meet the relevant harmonic grid codes, the OPP optimization problem considers the transfer function from the switching function harmonics to the grid current harmonics, while respecting the constraints that relate to the current harmonic limits. Moreover, the proposed problem relaxes the restrictions that are traditionally imposed on the switching signals, i.e., the quarter-wave symmetry and unipolar switch positions. In doing so, the current distortions are less than those of the conventional constrained OPPs, while they are occasionally even lower than those of the OPPs without constraints on the current harmonics. The presented numerical results demonstrate the superior harmonic performance of the computed OPPs compared with conventional modulation techniques, such as space vector modulation (SVM).

I. Introduction

Power electronic systems with power ratings of 1 MVA and above, such as medium-voltage (MV) converters, need to be operated at very low switching frequencies (well below 1 kHz) to keep the switching losses low. Conventional modulation methods, such as carrier-based pulse width modulation (CB-PWM) or space vector modulation (SVM), however, do not perform well at low pulse numbers, i.e., at a low ratio of switching-to-fundamental frequency [1]. To achieve high-quality converter outputs at low pulse numbers, programmed PWM schemes, such as selective harmonic elimination (SHE) and optimized pulse patterns (OPPs), can be employed instead [2], [3], in which the switching patterns are calculated in an offline procedure. Specifically, SHE computes the switching angles (i.e., switching time instants) of the switching patterns by solving a system of nonlinear equations such that specific harmonics are eliminated. As for the switching angles of the OPPs, these are computed by solving an optimization problem that accounts for the output current total demand distortion (TDD) in its objective function.