Solving Multi-Objective Optimization Problems using Differential Evolution Algorithm with Different Population Initialization Techniques

The researchers of Evolutionary Computing (EC) community proposing new and different algorithmic strategies to tackle the increasing issues in handling optimization problems. As the number of objectives in an optimization problem increases the algorithmic complexity in solving the problem also increases. The way the initial population for an optimization problem generated is greatly affecting the performance of the Evolutionary Algorithms (EAs). This paper investigates the performance of Differential Evolution (DE) in solving Mutli-Objective optimization problems (MOOP) with two different population initialization (PI) techniques. The performance of different instances of DE is compared based on the solution accuracy obtained. The results obtained shows that DE shows different performance for different PI techniques.