Optimal sizing of standalone hybrid energy system using a novel multi-objective evolutionary algorithm
Publication Date
February 22, 2020
Creator
Abstract
Many remote areas are subjected to lack of electricity access thus affecting a nation’s growth. Building standalone hybrid energy systems (HES) have been proven to be the most feasible solution. The inclusion of supercapacitor in HES aids in prolonging battery lifespan, reducing the overall system costs. To optimally size the system, genetic algorithm (GA) is initially chosen as the optimization tool with the first objective to minimize cost. Since real problems such as this often have multiple objectives that conflict one another, multi-objective evolutionary algorithm (MOEA) is preferred, which is a solver developed from GA itself. The second objective function is formulated to minimize unmet load, for which when the cost increases, the unmet load decreases, and vice versa. However, MOEA has its own conflicting goals in balancing between a trade-off’s convergence and diversity. Many MOEAs are designed to overcome the issue but each has their own pros and cons. Therefore, a new MOEA approach is proposed in this research to combine the convergence and diversity evaluations into one to eliminate goal bias. The experiments in this thesis include load profile generation for HES sizing, comparison between HES with and without supercapacitor, GA’s multiple runs to find minimum cost for fixed percentage of unmet load, the proposed MOEA’s test for functionality and robustness, and finally the comparison between GA’s and the proposed MOEA’s solutions to the HES optimal sizing problem. The results prove the supercapacitor’s advantage in reducing cost, the proposed MOEA’s reliability, and the unmet load minimization function’s practicability. This thesis concludes with summary and future work of this research.
Item Type
ethesis
Thesis Type
PhD
Supervisors
Subjects (LC)
Associated Schools / Departments
Department of Electrical and Electronic Engineering
eprints ID
59520
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
Hanim 016349 Revised Thesis.pdf
Type
Full-text
Description
Examined
Size
13.27 MB
Format
Adobe PDF
Checksum (MD5)
b7863c46433b17217272a91a3d334ca0