Sequential model predictive control of quasi Z-source inverter with fixed frequency operation

摘要

Quasi Z-source inverter (qZSI) is a single-stage high gain buck-boost converter suitable for various applications like renewable energy sources, power electronic drives, and power supply systems. Controller design for qZSI should ensure simultaneous control of multiple objectives such as the DC capacitor voltage, source inductor current, and load current. While controllers dealing with multiple objectives, obtaining a good steady-state and transient performance with low control complexity is essential. Fortunately, Model Predictive Control (MPC) is emerging as an effective alternative that can address all the above-said issues simultaneously. However, some of the severe problems to be addressed with the conventional MPC algorithms developed for qZSI are applying a single vector per sampling duration, higher computational requirements, and lack of analytic procedure to handle weight factors. This paper presents a control algorithm with features of weight factor elimination and AC load current ripple reduction by using a modified double vector predictive control approach with fewer computational requirements. The principle of inclusion of zero vector, weight factor elimination, and a detailed calculation for duty ratio are included. As the converter operates at a fixed frequency, significant ripple reduction for DC current, link voltage, and load current ripple have been achieved. The proposed MPC requires only approximate to 81% of computations needed for the conventional MPC. Methodological comparison has been provided to present the superiority of the proposed MPC compared to appropriate conventional MPCs available in the literature.