CQUniversity
Browse

Constrained multi-objective particle swarm optimization with application in power generation

Download (3.15 MB)
thesis
posted on 2021-10-31, 23:36 authored by Dujuan LiDujuan Li
This thesis is devoted to the study of metaheuristic optimization algorithms and their application in power generation. The study focuses on constrained multi-objective optimization using Particle Swarm Optimization algorithm. A multi-objective constraint-handling method incorporating a dynamic neighbourhood PSO algorithm is proposed for tackling single objective constrained optimization problems. The benchmark simulation results demonstrate the proposed approach is effective and efficient in finding the consistent quality solutions. Compared with the recent research results, the proposed approach is able to provide better or similar good results in most benchmark functions. The proposed performance-based dynamic neighbourhood topology has proved to be able to help make convergence faster than the static neighbourhood topology. The thesis also presents a modified PSO algorithm for solving multi-objective constrained optimization problems. Based on the constraint dominance concept, the proposed approach defines two sets of rules for determining the cognitive and social components of the PSO algorithm. Simulation results for the four numerical optimization problems demonstrate the proposed approach is effective. The proposed approach has a number of advantages such as being applicable to any number of objective functions and computationally inexpensive. As applications, three engineering design optimization problems and the power generation loading optimization problem are investigated. The simulation results for the engineering design optimization problems and the power generation loading optimization problem reveal the capability, effectiveness and efficiency of the proposed approaches. The methodology can be readily applicable to a broad range of applications.

History

Location

Central Queensland University

Open Access

  • Yes

Era Eligible

  • No

Supervisor

Professor Xinghuo Yu ; Dr. Wei Li

Thesis Type

  • Doctoral Thesis

Thesis Format

  • Traditional