Low-computational-complexity algorithm for current predictive control of an externally excited synchronous machine

Abstract

The current control in externally excited synchronous machines employed in applications such as automotive electrical traction drives is a challenging problem. These applications are characterized by fast dynamics that are subject to hard physical constraints including quadratic constraints related to the state and control input. The goal of this paper is to develop a controller synthesis method which can deal with these challenges. To this end, firstly, a polytopic approximation of quadratic constraints is performed. Then, a low-computationalcomplexity algorithm for current predictive control based on the flexible Lyapunov functions concept is employed by using a onestep-ahead prediction horizon. It is then shown that the resulting current control problem can be transformed into a linear program, which is suitable for an on-line implementation. A realistic simulation scenario is then used to illustrate the effectiveness of the proposed method. The obtained results show significant improvement compared with the existing PI approaches. Moreover, the computational complexity of the algorithm is compatible with hardware requirements of existing electronic control units.