Quadratic Regression Model Based Indirect Predictive Control of LC Filter-equipeed PMSM drives

Model Predictive Control has promising potential in controlling power electronic converters and electric drive systems. This approach promises reliable operation and multi-objective optimization even for nonlinear systems. In this paper the formulation of predictive control for a PMSM drive that has been equipped with an LC filter, which relieves the issues caused by the high dv/dt and di/dt that arise in converter fed electric machines. However, the complexity of the model in this case extends to six variables. Therefore, in the paper a regression model function is utilized that transforms the six-dimensional optimization problem into two dimensions. The solution to the optimization problem is straightforward to establish and the result can be forwarded to the modulation algorithm. The paper discusses the proposed predictive control method for the LC filter equipped PMSM drive, through experimental results for both steady-state and transient behaviours.