ABSTRACT

This chapter deals with the performance assessment of distinct Multi-objective optimization algorithms and the experimental validation of Pareto optimal solutions obtained thereof. Experiments were conducted on a phase change material-based 72 pin heat sink with 4 identical discrete heaters at the base. The goal of the experiments is to study the effect of discrete heating on the charging and discharging cycle. Researchers have been investigating the use of optimization techniques to perform thermo-geometric optimization of composite heat sinks. Different combinations of heat inputs are generated using the Latin Hypercube Sampling technique. Multi-objective particle swarm optimization was developed first in 1995, inspired by an animal social behaviour simulation system that incorporated concepts such as nearest-neighbour velocity matching and acceleration by distance. K-means clustering is a well known simple non-supervised clustering algorithm widely applied in the solutions of multi-objective optimization algorithms.