Optimal synthesis and design of dynamic systems under uncertainty

Abstract A unified process synthesis framework for obtaining process designs and control systems, which are economically optimal while being able to cope with process variations, is presented in this paper. Process systems are modelled via dynamic mathematical models (DAE's), variations include both uncertain parameters and time-varying disturbances, while control structure selection and controller design is considered as part of the optimization problem. Robust stability criteria are also included within the synthesis scheme so as to maintain desired dynamic characteristics. An efficient decomposition algorith for the solution of the resulting mixed-integer stochastic optimal control problem is outlined and demonstrated with a ternary distillation example.