Browse By Repository:

 
 
 
   

Incorporating space division at initialization phase of bees algorithm

Chelliah, Ratnadevi (2017) Incorporating space division at initialization phase of bees algorithm. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

[img] Text (24 Pages)
Incorporating Space Division At Initialization Phase Of Bees Algorithm - Ratnadevi Chelliah - 24 pages.pdf - Submitted Version

Download (308kB)
[img] Text (Full Text)
Incorporating space division at initialization phase of bees algorithm.pdf - Submitted Version
Restricted to Repository staff only

Download (1MB)

Abstract

Bees Algorithm is a population-based optimization algorithm that mimics the natural foraging behavior of swarm bees. At initialization phase in Bees Algorithm, uniform random distribution is used where the randomness help the system to sample new solutions particularly in complex optimization problems such as in computer science, engineering control and mechanical design. Unfortunately this leads to slow convergence rate to the optimum solution in Bees Algorithm. Therefore, in order to increase the convergence speed, Search Space Division (SSD) is implemented at initialization phase of the existing Bees Algorithm. Search Space Division is a novel initialization method that can easily work on high dimension. The trustworthiness of new improved algorithm is evaluated by comparing success rate with Bees Algorithm and quick Artificial Bee Colony. Success rate of Search Space Division Bees Algorithm outperformed both Bees Algorithm and quick Artificial Bee Colony. The efficiency of the new improved method in terms of convergence speed is compared with standard Bees Algorithm. Both algorithms were tested by using different parameter settings on selected benchmark test functions. Results show that all four parameters have equally good results on improved algorithm. According to comparison of mean evaluation of both algorithms shows significant improvement of new algorithm in convergence speed than existing Bees Algorithm. However, t-test shows newly improved convergence speed is on par with standard Bees Algorithm. Thus, incorporating Search Space Division in initialization phase of Bees Algorithm serves as a preliminary work to reveal the convergence speed of the newly improved algorithm.

Item Type: Final Year Project (Project Report)
Uncontrolled Keywords: Engineering design, Artificial intelligence, Algorithms, Engineering mathematics
Subjects: T Technology > T Technology (General)
T Technology > TA Engineering (General). Civil engineering (General)
Divisions: Library > Final Year Project > FKP
Depositing User: Nor Aini Md. Jali
Date Deposited: 10 Dec 2018 02:46
Last Modified: 14 Nov 2023 07:29
URI: http://digitalcollection.utem.edu.my/id/eprint/22188

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year