群知能と進化的計算の国際ジャーナル

概要

Hybridization of Fruit Fly Optimization Algorithm and Firefly Algorithm for Solving Nonlinear Programming Problems

Rizk M Rizk Allah

We propose a novel hybrid algorithm named, FOA-FA to solve the nonlinear programming problems (NLPPs). The main feature of the hybrid algorithm is to integrate the strength of fruit fly optimization algorithm (FOA) in handling continuous optimization and the merit of firefly algorithm (FA) in achieving robust exploration. The methodology of the proposed algorithm consists of two phases. The first one employs a variation on original FOA employing a new adaptive radius mechanism (ARM) for exploring the whole scope around the fruit flies locations to overcome the drawbacks of original FOA which has been continues for the nonnegative orthant problems. The second one incorporates FA to update the previous best locations of fruit flies to avoid the premature convergence. The hybrid algorithm speeds up the convergence and improves the algorithm’s performance. The proposed FOA-FA algorithm is tested on several benchmark problems and two engineering applications. The numerical comparisons have demonstrated its effectiveness and efficiency.