Pipelining Strategies and Design Considerations of Predictive Current Control Method

Abstract

This paper explores the pipelining strategies for the model predictive control methods. The array and vector processing methods are examined to discover their applicability in the model predictive current method. The potential benefits of the pipelining methods are investigated, and their design methodologies are scrutinized. The model predictive control is a nonlinear control technique that predicts the system dynamics. The model predictive control (MPC) provides rapid response to the load variations and guarantees robust operation. However, the lower sampling period is the main design constraint to achieve a reliable system operation. The selection of a low sampling period demands a powerful digital controller due to the increasing computational burden. To handle the high calculation burden, a field-programmable gate array (FPGA) is a powerful solution. A proper pipelining strategy enables the use of the MPC in real-time applications. In this paper, pipelining strategies and practical design considerations of the FPGA-based predictive current method are presented. The nine switch converter (NSC) is selected as an experimental case study. The experimental results are provided to demonstrate the theoretical framework. The experimental results prove the feasibility of the array processing and vector processing methods in MPC applications.