Application of Nondominated Sorting Genetic Algorithm for Multiobjective Optimal Design of Distribution Transformer

Abstract

In today’s competitive environment, transformer manufacturers are faced with the exigent task of yielding optimum performance at lowest cost. Considering the aspect of energy shortage and increase in its cost, the complexity of achieving the optimal balance between transformer manufacturing costs and transformer performance is a herculean task, demanding great amount of attempts to reach satisfactory results. To cater with the problem of minimizing transformer losses and cost simultaneously, this paper deals with multiobjective optimization of distribution transformer using binary coded nondominated sorting genetic algorithm (NSGA-II). The design procedure takes into account three objectives: active part cost, no-load losses and load losses of a distribution transformer. Elitist nondominated sorting and crowding distance are used to obtain pareto optimal solutions. Results indicate the potential of NSGA-II in maintaining diversity among solutions. To enable the decision maker (DM) to make a choice between different pareto-optimal solutions, TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) technique is then suggested for obtaining best compromise solution among nondominated solutions. The effectiveness of the proposed method has been demonstrated on 100 kVA distribution transformer.

Keywords: Multiobjective optimal transformer design, Pareto-optimal solutions, NSGA-II, TOPSIS

Cite this Article

Mehta HD, Rajesh Patel. Application of Nondominated Sorting Genetic Algorithm. Trends in Electrical Engineering. 2016; 6(3): 44–57p.