Predictive control of a permanent magnet synchronous generator connected to an MMC converter in an oscillating water column based power plant

Abstract

Modular Multilevel Converters (MMC) are traditionally used in high-power applications although their field of application has lately widened to lower powers. In medium and low-power applications such as electric drives, a fast control system is usually required and model-based predictive control approaches (MPC) are highly suitable to obtain superior performance. Unfortunately, although MPC is widely used with two-level three-phase power converters, typically in motor and generator control applications, it has been barely developed for MMC. Recent studies discuss MPC approaches for MMC but they are primarily focused on improving the internal operation of MMC. In contrast, this paper presents an MPC which is designed to control the torque and the magnetic field of a surface permanent magnet synchronous generator (SPMSG) by means of an MMC. The proposed approach features a new and fast method to find the location of the reference voltage vector among the large number of multilevel vectors present in typical MMCs. The performance of the new control system has been demonstrated for an oscillating water column (OWC) based power plant simulation. The control system stands out for its fast response, low THD currents and reasonably low computing time which enables its easy implementation in microcontrollers (MCU).