Convergence Analysis of the African Buffalo Optimization Algorithm

Abstract

This paper presents the convergence analysis of the newly-developed African Buffalo Optimization algorithm. African Buffalo Optimization is a simulation of the organizational skills of the African buffalos using two basic sounds: /waaa/ and /maaa/ as they transverse the African landscape in search of grazing pastures. The African Buffalo Optimization has proven to be quite successful since its development hence the need to examine its convergence behaviour. The analysis of the convergence of Natureinspired optimization algorithms is necessary to help researchers and practitioners understand the workings of the algorithms in the algorithms' attempts at solutions. After a number of evaluations, this study discovered that the convergence of African Buffalo Optimization is a function of the population size, communication topology, parameter-set, landscape topology and the objective function being optimized.

African Buffalo Optimization for Global Optimization

“…A closer observation, however, reveals that the above four benchmark functions represent three unconstrained and one constrained function. This outcome is in agreement with earlier findings that the constrained function requires a longer procedure for optimization, because it may have to be first converted to unconstrained function [33][34][35] . Conversely, all the algorithms had their worst performance in the egg-holder function (Figure 7).…”

Transmission network expansion static planning considering security constraints via African Buffalo Algorithm

“…The ABO(N-1) optimizer was implemented in MatLab language and simulations were performed on a computer with an Intel Core i5-7400T, 2.40 GHz, 8 GB RAM processor. The data used to adjust buffalo displacement velocities were lp1=0.9 and lp2=0.7 suggested in [14] to obtain a better balance between intensification and diversification processes of search space exploration.…”

“…In this step the data for the ABO algorithm is provided, i.e., the learning factors lp1 and lp2 used in [14] to adjust the velocities of buffalo displacements from the pasture regions in search of food.…”

“…As can be observed, the application of the metaheuristic algorithm African Buffalo Optimization -ABO [13], [14] to solve the planning problem, both in the static and dynamic versions, with or without safety constraints, is not included in the list of published articles. This novelty, coupled with the fact that the discussion of the TNESPS subject is still open, was what motivated this work.…”

Engenharia de Produção: paradigmas e aplicabilidades no setor produtivo

“…28 , i.e., 526 times larger than the number of possible combinations of additionsto the G-6/15 system. The global optimal solution that the ABO (N-1) optimizer found, without load shedding, contains 28 circuits, added on 17 existing branches (n 01-05 =2, n 03-09 =1, n 03-24 =2, n 04-09 =1, n 05-10 =1, n 06-10 =2, n 07-08 =3, n 10-11 =1, n 11-13 =1, n[14][15][16] =2, n 15-16=1 , n 15-21 =1, n 15-24 =2, n 16-17 =3, n 16-19 =2, n 17-18 =2 and n 21-22 =1) and costs $1,071 million.…”