(139a) Economic Model Predictive Control With Switching Objective Function and Constraints: Merging Actuator Preventive Maintenance and Economically-Optimal Operation

Next-generation advanced manufacturing is emerging as a necessity for the U.S. chemical process industry to remain competitive in the global markets. Essential to the objectives of advanced manufacturing is development of preventive maintenance programs for control system components like actuators and sensors. A high percentage of the small-scale, day-to-day preventive maintenance tasks for the chemical process industry is for control actuators of process control systems (e.g., pump rebuilds, valve replacements, etc.). To accomplish these preventive maintenance tasks, the ability for a control system to maintain stable operation of the process while operating the process in an economically optimal manner with respect to the available control actuators is desirable. One natural framework to accomplish this task is to use fault-tolerant control which has been extensively researched [1] in the context of reactive fault-tolerant control. Specifically, a new emerging area within fault-tolerant chemical process control is proactive fault-tolerant control which deals with control systems that take proactive action to actively reconfigure the control system and to compensate for changing number of control actuators which allows preventive maintenance to be completed on the control actuators [2].

The focus of this work is to develop a proactive fault-tolerant Lyapunov-based economic model predictive control (LEMPC) method that can deal with changing number of manipulated inputs as a result of control actuators being taken out of service for preventive maintenance or placed back into service after the maintenance work has been completed. LEMPC is an attractive choice for proactive fault-tolerant control because it is able to determine economically-optimal, time-varying operating trajectories while account for process constraints [3]. However, LEMPC cannot be applied directly because the optimization problem dimensionality, cost function, and constraints change as a result of the changing number of inputs. Furthermore, the proactive LEMPC must proactively account for the changing number of manipulated inputs such that the stability of the process can be maintained after the number of manipulated inputs changes (which results in modification of the closed-loop stability regions). To accomplish these goals, we formulate and prove stability for a proactive fault-tolerant LEMPC and demonstrate through simulations of a realistic chemical process under the proactive fault-tolerant LEMPC that the LEMPC is able to maintain stability of the process, perform successful reconfiguration of the control system accounting for variable number of manipulated inputs, and optimally operate the process.

References

1. Mhaskar P, Liu J, Christofides PD. Fault-Tolerant Process Control: Methods and Applications. London, England: Springer-Verlag, 2013.
2. Lao L, Ellis M, Christofides PD. Proactive fault-tolerant model predictive control. AIChE Journal, in press.
3. Heidarinejad M, Liu J, Christofides PD. Economic model predictive control of nonlinear process systems using Lyapunov techniques. AIChE Journal. 2012;58:855-870.