Algorithmic synthesis and integrated design of chemical reactor systems using genetic algorithms

This work presents an algorithmic approach to allow for the synthesis and the integrated design of processes considering simultaneously economics and controllability. The method is applied to the synthesis and design of a continuous-stirred-tank-reactor (CSTR) system. For the synthesis. all the design alternatives are coded into a superstructure which translates into a mixed integer nonlinear (MINLP/DAE) optimization problem. The solution gives the best process configuration, together with dimensions and optimal operating point. In the integrated design a compromise solution among economics and control objectives is determined by solving a multiobjective constrained non-linear optimization problem. Both problems are solved using genetic algorithms.