High-performance Embedded Model Predictive Control using Step Response Models

Abstract This paper presents an efficient approach used to obtain high-performance control targets for an embedded MPC scheme, which has a relatively low computational complexity compared to a field proven industrial MPC framework. The results show that prediction quality is an essential factor to consider when using the embedded MPC scheme for challenging industrial applications where step response models are used. A recursive prediction model implementation and a more straightforward implementation, referred to as the standard implementation are examined in this paper. The results show that both implementations have inherent properties that directly affect control performance in the presence of disturbances. The recursive implementation turns out as the best choice for the embedded MPC due to its superior robustness and computational properties. The impact of the prediction model implementation on the control of a subsea compact separation process is also studied.