Duty-Cycle-Optimized Voltage Predictive Control for Four-Leg Inverters With Over-Modulation Capability

Three-phase four-leg inverters are effective solutions to solve three-phase uncertain loads problem. However, the irregular three-dimensional vector space of four-leg inverter makes it difficult to achieve high steady-state performance and ripple reduction. To overcome this deficiency, a duty-cycle-optimized voltage predictive control method (DCO-VPC) is proposed to improve the performance of conventional modulated model predictive control (C-M2PC) including output ripple reduction. It maintains a fixed switching frequency meanwhile inheriting the fast dynamic response and over-modulation capability of finite control set model predictive control (FCS-MPC). Virtual vectors are generated by angular errors between vectors, which partition the spatial tetrahedron into finer subspace and iterative operation is performed to improve prediction accuracy. In addition, a Lagrange function with Karush-Kuhn-Tucker (KKT) conditions is constructed and solved to obtain the output voltage boundary of inverter in the direction of current reference vector, which is used as a criterion to determine over-modulation region of four-leg inverter. Comprehensive experiments with different control schemes under various operating conditions show that the proposed DCO-VPC has advanced steady-state and transient performance, as well as over-modulation capability.