DEVELOPMENT OF A MULTI-OBJECTIVE OPTIMIZATION FRAMEWORK FOR INDUSTRIAL PROCESSES

• Main Author: GANESAN, TIMOTHY
• Format: Thesis
• Language: English
• Published: 2013
• Subjects: TP Chemical technology
• Online Access: http://utpedia.utp.edu.my/22105/1/2013%20-CHEMICAL%20-%20DEVELOPMENT%20OF%20A%20MULTI-OBJECTIVE%20OPTIMIZATION%20FRAMEWORK%20FOR%20INDUSTRIAL%20PROCESSES%20-%20TIMOTHY%20GANESAN.pdf; http://utpedia.utp.edu.my/22105/

Multi objective (MO) optimization is an emerging field which is increasingly being applied in many engineering-based industries globally. Whilst MO optimization problems have been actively researched for the past years, industrial-scale problems involving objectives more than two are rarely studied although frequently encountered. Thus, this work focuses on industrial MO problems encountered in process engineering with more than two objectives. This work investigates the performance and solution characteristics of metaheuristic algorithms when applied to such problems. In addition, the solution characteristics produced by these algorithms were also analysed and its influence on the algorithmic performance was ascertained. The knowledge obtained from the above explorations was then used to construct a novel framework for solving these problems efficiently. The proposed framework is introduced as the 'surgery framework (SF)' since its prime functionality is to dissect the algorithm and add certain mechanisms into it. The SF introduced here comprises of the measuring the solutions of multiple algorithms as well as modifying them by adding certain mechanisms into chosen algorithms. The SF was then used to solve four real-world industrial MO process optimization problems. The industrial problems are as follows; Application I: MO Optimization of Extraction Process of Bioactive Compounds from Gardenia, Application II: MO Optimization of Bioethanol Production during Cold Enzyme Starch Hydrolysis, Application Ill: MO Optimization of Synthesis Gas for Production Process and Application IV: MO Optimization of Pre-treatment Strategy for Bioethanol Production from Rice Husk. The SF applies four metaheuristic algorithms such as differential evolution (DE), genetic algorithm (GA), gravitational search algorithm (GSA) and particle swarm optimization (PSO) to solve each of the industrial MO process optimization problems.