The Williams and Otto Chemical Plant re-evaluated using a Pareto-optimal formulation aided by Genetic Algorithms

Abstract

The Williams and Otto Chemical Plant is a classic example of a complex nonlinear programming problem incorporating the essential features of a chemical or hydrometallurgical processing plant. An efficient solution strategy for this time tested problem is shown here which uses different variants of biologically inspired Genetic Algorithms and is aided by a multi-objective formulation. The Genetic Algorithms worked more efficiently than the classical techniques used earlier and the nature of the feasible solutions was clearly revealed by the Pareto-optimality concept embedded in the multi-objective optimization employed here.