Application of NSGA-II Algorithm to Single-Objective Transmission Constrained Generation Expansion Planning

Abstract

This paper presents an application of elitist nondominated sorting genetic algorithm version II (NSGA-II), a multiobjective algorithm to a constrained single objective optimization problem, the transmission constrained generation expansion planning (TC-GEP) problem. The TC-GEP problem is a large scale and challenging problem for the decision makers (to decide upon site, capacity, type of fuel, etc.) as there exist a large number of combinations. Normally the TC-GEP problem has an objective and a set of constraints. To use NSGA-II, the problem is treated as a two-objective problem. The first objective is the minimization of cost and the second objective is to minimize the sum of normalized soft constraints violation. The hard constraints (must satisfy constraints) are treated as constraints only. To improve the performance of the NSGA-II, two modifications are proposed. In problem formulation the modification is virtual mapping procedure (VMP), and in NSGA-II algorithm, controlled elitism is introduced. The NSGA-II is applied to solve TC-GEP problem for modified IEEE 30-bus test system for a planning horizon of six years. The results obtained by NSGA-II are compared and validated against single-objective genetic algorithm and dynamic programming. The effectiveness of each proposed approach has also been discussed in detail.