High Performance Modified Model Predictive Control of a Voltage Source Inverter.

This paper proposes a modified model predictive control (MMPC) method for a two-level voltage source inverter (VSI) by applying 'Midpoint Euler's Approximation' technique to achieve a high performance. Therefore, a new predictive mathematical modeling is derived and presented in this paper to predict the future behavior of the inductive load model, which is compared with the reference signal to determine the cost function of the system. In this MMPC method, all the possible switching vectors of the converter are utilized to obtain the cost functions, and the corresponding switching vector for minimum cost function is selected to actuate the power converter in the next sampling time interval. The proposed MMPC method and the traditional MPC schemes are verified in MATLABSimulink. The experimental validation is performed in DS1104 R&D Controller Board to justify and confirm the performance of the system. Simulation and experimental results validate the robustness of the proposed MMPC method, and improve the results compared to traditional MPC method in terms of total harmonic distortion (THD) and current reference tracking mean error. [ABSTRACT FROM AUTHOR]