Adaptive Control of a Continuous Polymerization Reactor

In this study, an adaptive control strategy is applied to a digital simulation describing the continuous polymerization of MMA. The reactor model illustrates unusual process dynamics such as multiple steady states, limit cycles, and steady state instabibilities. This dissipative behavior results from the autocatalytic nature of polymerization reactions owing primarily to the "gel effect." The MIMO self-tuning regulator of Vogel and Edgar is applied to this model. The controller is based upon pole-zero placement and includes dead time compensation for variable and multiple delays. Parameter estimation for the controller is performed by recursive least squares. Initiator concentration in the feed and the reactor jacket temperature are used effectively to control the monomer concentration and temperature in the reactor.