AN INDUSTRIAL APPROACH FOR EFFICIENT MODELING AND ADVANCED CONTROL OF CHEMICAL BATCH PROCESSES

Chemical batch processes offer an attractive way of producing a variety of specialty products in a highly flexible manner. However, such processes are hard to control due to the absence of the notion of a steady state operation - necessitating the use of nonlinear models- and the fact that product qualities are only measured at the end of each batch. This paper proposes a new industrial modeling and control strategy that allows significant batch time reductions to be obtained, taking physical, safety and quality constraints into account. Application to an industrial reactor shows significant improvements over classical control strategies.