Hierarchical Economic NMPC for a Class of Hybrid Systems Using Neighboring-Extremal Updates

摘要

A two-layer control algorithm is developed for a class of hybrid (discrete-continuous dynamic) systems comprising important applications such as the economically optimal operation of recipe-driven batch or continuous processes. On the upper layer, the economic optimal control problem is solved rigorously by a slow controller at a low sampling rate, whereas a fast neighboring-extremal controller updates rather than tracks the optimal trajectories to account for disturbances. Consequently, the process is steered to its operational bounds, while the optimal switching times under disturbances can be determined such that the economic potential of the process is fully exploited anytime.