Studies of evolutionary algorithms and applications in reliability optimization

Abstract

In this thesis, we have investigated different types of reliability optimization problems that have been formulated and solved as single and multi-objective constrained optimization problems with integer and/or mixed-integer variables using different hybrid algorithms. Chapter 2 makes an overview of past and recent developments on different evolutionary algorithms. In this chapter, the discussion is mainly focused on a literature survey about Evolutionary Algorithms (EA), Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Differential Evolution (DE), Ant Colony Optimization (ACO), Reliability Optimization Problem (ROP), Wireless Sensor Network (WSN) and Power Distribution System (PDS), GA-PSO and GA-ACO hybrid algorithms. Chapter 3 deals with an overview of existing finite interval mathematics, and fuzzy sets, defuzzification processes. In this chapter, we have also discussed about GA PSO, DE, ACO, GA-PSO and GA-ACO hybrid algorithms. Chapter 4 deals with the development of an efficient hybrid GA-PSO approach based on genetic algorithm and particle swarm optimization for solving mixed integer nonlinear reliability optimization problems in series, series-parallel and bridge systems. This approach maximizes the overall system reliability subject to the nonlinear resource constraints arising on system cost, volume and weight. Chapter 5 presents multi-objective reliability-redundancy allocation problem by hybrid optimization techniques. This technique is based on combination of genetic algorithm and particle swarm optimization. In this chapter, we have solved mixed integer nonlinear multi-objective reliability optimization problem using hybrid GA-PSO algorithm. Chapter 6 focuses on Reliability Redundancy Allocation Problem (RRAP) in Wireless Sensor Network (WSN) system is obviously an important problem. In this chapter, a decision making assessment of reliability of Redundancy Allocation Problem (RAP) is proposed using fuzzy approach. Chapter 7 represents the formalization of an optimization problem that jointly minimizes the afore-mentioned reliability indices as well as the cost of a PDS by optimal allocation of different protective devices and switches have always been a challenging task. This chapter presents a hybrid single as well as joint-objective function optimization using hybrid GA-ACO algorithm.