Abstract:
Model predictive control (MPC) has been broadly used in process industry for tracking pre-determined set points, which are associated with optimal economic performance. This implies that we require to first obtain a desired set-point of the key process parameter through process design with the help of static optimization. This process parameter can then be tracked with the help of online optimization in a standard MPC. On the other hand, Economic MPC (EMPC) makes use of an economic cost function in the optimization problem solved at every sampling instant in MPC to directly achieve a desired objective without having to pre-determine a set-point. This EMPC approach relaxes the assumption of process being operated under steady state conditions. In fact, EMPC approach leads to process operation under forceful un-steady state conditions. This in turn is supposed to enhance the process efficiency since a relaxed optimization problem is solved unlike in the standard MPC. Benefits of EMPC over standard MPC are demonstrated in a non-isothermal continuous stirred tank reactor with single and multiple objectives.
Bio-sketch: Dr. Nitin Padhiyar is an Associate Professor in Chemical Engineering at IIT Gandhinagar. He did his BTech, MTech, and PhD from Gujarat University, IIT Kharagpur, and IIT Bombay. Dr. Nitin’s research work focuses on optimization, multi-objective optimization and model predictive control with single and multi-objective optimization. He is the recipient of the Ambuja Young Scientist Award and Indo-US Research Fellowship.