The development and application of a multi-objective optimization technique for chemical processes and controller design

Abstract

In recent years, the development and application of multi-objective optimization techniques have received increasing attention in the literature. Recent innovations in the fields of systems science, artificial intelligence, and operations research have led to the development of rigorous multi-objective optimization techniques to address the problem of optimizing complex processes in the presence of multiple conflicting objectives. This thesis is a collection of three papers that focuses on the development and application of a multi-objective optimization strategy for selecting optimum operating conditions for the production of gluconic acid and for determining optimum tuning parameters for PID controllers. The optimization strategy developed is performed in two steps: the approximation of the set of feasible solutions called the Pareto domain using the Dual Population Evolutionary Algorithm and the classification of the domain using the Net Flow Method, which incorporates information on the process provided by an expert. This strategy has been proven to be robust in determining the optimal solution after studying twelve standard test cases, which have been used frequently in the literature, and two engineering problems. In addition, the Pareto domain per se provides very useful information on the quality of the optimal zone.