Computationally efficient multi‐step direct predictive torque control for surface‐mounted permanent magnet synchronous motor
Changliang Xia,Zhanqing Zhou,Zhiqiang Wang,Yan Yan,Tingna Shi
DOI: https://doi.org/10.1049/iet-epa.2016.0221
2017-05-01
IET Electric Power Applications
Abstract:The multi‐step direct model predictive control (M‐DMPC) could reduce the switching frequency and improve the efficiency of the motor drive system while ensuring good steady and dynamic performance. However, the conventional M‐DMPC, which uses the exhaustive optimisation method is with exponential time complexity, so it is difficult to realise in a short sample period. This study proposes a multi‐step direct predictive torque control (M‐DPTC) algorithm for two‐level voltage source inverter fed surface‐mounted permanent magnet synchronous motor (SPMSM) drive system. In the proposed algorithm, first, in order to reduce computation time in multi‐step predictive process, a simplified multi‐step predictive model is established. Compared with the complex operations, like square root and trigonometric function used in the conventional predictive model, only look‐up table and addition operation are utilised in the proposed model for multistep prediction. Second, compared with the exhaustive searching method, the proposed optimisation method avoids the exploration of all possible switching sequences and has logarithmic time complexity. Combining the above two measures, the proposed M‐DPTC can be achieved in a relatively shorter sample period. Simulation and experimental results for a 6‐kW SPMSM are presented to validate the proposed algorithm.
engineering, electrical & electronic