A Suggested Approach to Practical Unclamping of Model Predictive Control

Model predictive control is a form of advanced control that is widely used in industry. One of the reasons why it is so common is because of the ease with which control and state constraints can be included into the control formulation. Correct setting of the manipulated variable constraints (limits) are important for control stability and optimality. This paper provides a suggestive approach to solving the problem of process plant operators setting manipulated variable limits to sub-optimal values (referred to in this paper as clamping). The optimal values are calculated through solving a constrained optimization problem, and these values are reported to the operator, who may decide to move the variable limit (s). This approach helps to further optimal operation while not taking control of the limits away from the plant operator. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.