An approach of constraint boundaries tuning based on shadow price for two-layered predictive control

Shadow prices show the effect, which were caused by variations of constrained boundaries, to optimum value of objective function under the current optimal strategy. In this paper, a LP-MPC form of two-layered predictive control was described, and a constraint tuning strategy based on shadow price was proposed under this structure. The nature of disturbance can be evaluated according to the history data of process, then, combine with constraint conditions of the process, tuning boundaries will be obtained. The shadow prices for constrained boundaries of steady-state target calculation were counted based on solving a linear programming and its dual problem. Then, the constraint boundaries, which influence the objective optimum effectively, were handled selectively. The process will be pushed to allowable operation boundaries, to increase the economic benefit. Finally, in a practical process, a simulation example was conducted in order to verify the useful of shadow price to constraint tuning for two-layered predictive control.