Development of mixed constrained RTDA controller for industrial applications

Most of the industrial processes are MIMO and constrained in nature. So the development of MIMO constrained controller is always preferable and makes challenges to the control community. This paper proposes the development of constrained next generation RTDA controller to handle various linear inequality constraints in a multi-variable control framework. The Lagrangian function is formulated, and the dual problem is solved using Hildreth's algorithm. The combined hard-soft constrained RTDA optimization problem with exact penalty functions is solved using KKT conditions. Performance of the developed constrained RTDA controller is demonstrated using simulation studies of typical industrial unit operations such as nonlinear CSTR process and Wood and Berry distillation column. The performance of the constrained RTDA controller is compared with constrained DMC and GPC.