Dual-stage model predictive control of modular multilevel converter

Abstract

The control of modular multilevel converter (MMC) is quite challenging and demands a flexible approach to achieve multiple control objectives. The model predictive control (MPC) is highly effective and suitable to control the MMC. The control objectives are included in a single cost function and evaluated for all possible switching states. The switching state which minimizes cost function is selected and applied to converter. In MMC, the number of available switching states are quite high and drastically increase with number of submodules per arm. Therefore, the implementation of conventional MPC is difficult and contributes to unwanted switching of submodules. This paper proposes a less computational, dual-stage MPC with common-mode voltage injection for MMC. In this approach, the overall MMC control objectives are categorized into primary and secondary group. The primary group objectives are evaluated in first-stage and secondary group of control objectives are evaluated in second-stage of MPC. By doing so, the computational complexity can be significantly minimized without affecting the dynamic response of MPC. The proposed approach avoids unwanted switching of submodules and, minimizes the output voltage harmonic distortion and ripple in output current. The simulation study is conducted to verify the effectiveness of proposed approach and corresponding results are compared with standard MPC.