A New Distributed Evolutionary Computation Technique for Multi-Objective Optimization Md. Asadul Islam 1 , G.M. Mashrur-E-Elahi 2 , M.M.A. Hashem 3 Department of Computer Science and Engineering Khulna University of Engineering & Technology (KUET) Khulna 9203, Bangladesh asad_kuet@yahoo.com 1 , ranju2k4cse_kuet@yahoo.com 2 , mma_hashem@hotmail.com 3 Abstract— Now-a-days, it is important to find out solutions of Multi-Objective Optimization Problems (MOPs). Evolutionary Strategy helps to solve such real world problems efficiently and quickly. But sequential Evolutionary Algorithms (EAs) require an enormous computation power to solve such problems and it takes much time to solve large problems. To enhance the performance for solving this type of problems, this paper presents a new Distributed Novel Evolutionary Strategy Algorithm (DNESA) for Multi-Objective Optimization. The proposed DNESA applies the divide-and-conquer approach to decompose population into smaller sub-population and involves multiple solutions in the form of cooperative sub-populations. In DNESA, the server distributes the total computation load to all associate clients and simulation results show that the time for solving large problems is much less than sequential EAs. Also DNESA shows better performance in convergence test when compared with other three well-known EAs. Keywords- Novel evolutionary algorithm; time variant mutation; subpopulation; MOPs; Distributed computing; I. INTRODUCTION Multi-Objective optimization optimize a set of conflicting objectives simultaneously. MOP is a very important research topic, not only for the Multi-Objective nature of most real- world decision problems, but also there are still many open questions in this area. Traditionally, there are several methods available in the Operational Research (OR) literature for solving MOPs as mathematical programming models [1]. None of the OR methods treats all the objectives simultaneously which is a basic requirement in most MOPs. In addition, these methods handle MOPs with a set of impractical assumptions such as linearity and convexity. In MOPs, there is no single optimal solution, but contains a set of alternative solutions. These solutions are optimal in the wider sense since there is no other solutions in the search space that is superior to them when all objectives are simultaneously considered. They are known as pareto-optimal solutions or non-dominated solutions [2][3]. In principle, multiobjective optimization is difficult than the single-objective optimization. In single objective optimization, one attempts to obtain the best design or decision, which is usually the global minimum or the global maximum depending on the optimization problem. Recently, EAs are found to be useful for solving MOPs [4]. EAs have some advantages over traditional OR techniques. This allows us to find several members of the Pareto-optimal set in a single run of the algorithm [5]. Also, there is no requirement for differentiability of the objective functions and the constraints. Moreover, evolutionary algorithms are susceptible to the shape of the Pareto front and can easily deal with discontinuous or concave Pareto front. There is no well- accepted method for MOPs that will produce a good set of solutions for all problems. Also, sequential EAs require an enormous computing power and take much time to solve large problems. This motivates the further development of good approaches to MOPs. This paper proposes a parallel evolutionary algorithm called Distributed Novel Evolutionary Strategy Algorithm (DNESA) for MOPs, which is developed from a sequential evolutionary algorithm called Novel Evolutionary Strategy (NES) algorithm [10]. The proposed DNESA uses parallel approach to find out the member of pareto-optimal solution, which is more promising than sequential approach. Also diversity of population and fast convergence are achieved by using subpopulation and parallelization. In DNESA, the server divides the problem loads into all associate clients and the simulated results show that it takes less time to solve large problems than sequential EAs. For the convergence test, DNESA is compared with three algorithms namely Pareto Enveloped-based Selection Algorithm (PESA) [6], Non- Dominated Sorting Genetic Algorithm (NSGA-II) [7] and the Strength Pareto Evolutionary Algorithm (SPEA-2) [8]. The compared results showed that DNESA achieved better convergence with respect to other three algorithms. The organization of the paper is as follows. Section II discusses various existing techniques of Multi-Objective optimization. A brief discussion about the proposed approach is presented in section III. Experimental setup and results is presented in section IV. Section V concludes the paper. II. BACKGROUND INFORMATION A. Problem Definition The MOPs [9] (also called multicriteria optimization, multiperformance or vector optimization problem) can be