Optimal Batch Process Regulation using Self-optimizing Control, NCO tracking

Optimal operation of batch processes in the presence of uncertainty has been an area of significant research interest. The use of online measurements as a source of information offers scope to mitigate the effects of uncertainty and attain optimal operation. Therefore, amongst the different approaches proposed to address the uncertainty issues, measurement based optimization approaches have been shown to be eminently suited. Measurements have been broadly used in two approaches, viz. self optimizing control (SOC) and tracking of necessary conditions of optimality (NCO). In an earlier work for continuous processes by Jäschke and Skogestad [1], the relative complementary roles of NCO tracking and SOC were clearly highlighted in terms of their applicability in combating the effects of different scales of uncertainty and disturbances. For batch processes, the NCO tracking is found to be successful, the complementary role of SOC is largely un-explored. In this work, we extend the integrated approach of Jäschke and Skogestad [1] to batch processes by redefining the NCO and SOC layers. The proposed framework is evaluated using simulations involving a nonlinear batch polymerization of styrene.