On the robust application of stochastic optimisation technology for the synthesis of reaction/separation systems

Stochastic optimisation technology is presented for the synthesis of integrated reaction and separation process superstructures. The optimisation techniques enable the systematic exploration of process design options hidden in the network formulations. The implementation of two stochastic algorithms, simulated annealing and Tabu search, is presented in the context of the synthesis problem. The paper presents a design transition framework at the interface between the optimisation technology and the process synthesis representation as well as customisation aspects that enable an efficient search of the vast number of processing candidates generally comprising the solution space of the optimisation problem. Both stochastic algorithms are easily implemented as they utilise the same state transition and simulation framework. Asymptotic convergence to the globally optimal domain is observed for simulated annealing as well as Tabu search. In view of the numerical evidence obtained, Tabu search appears to arrive at high-quality solutions on more efficient search paths as compared with simulated annealing. (C) 2002 Elsevier Science Ireland Ltd. All rights reserved.