The IABC adopts the reward-based roulette wheel selection mechanism initially to divide all solutions suitably into feasible and infeasible solutions; thereafter, it divides them based on feasible and infeasible solutions for the implementation of incentives and punishments. Finally, the proposed method is applied to nonlinear system control problems. | Journal of Automation and Control Engineering Vol. 4, No. 1, February 2016 Approximation of the Piecewise Function Using Neural Fuzzy Networks with an Improved Artificial Bee Colony Algorithm Cheng-Hung Chen, Yao-Cheng Tsai, and Rong-Zuo Jhang Department of Electrical Engineering, National Formosa University, Taiwan Email: numerical functions, Karaboga introduced a bee swarm algorithm called the artificial bee colony (ABC) algorithm that simulates the foraging behavior of bees [10]. This study presents an improved artificial bee colony (IABC) for NFNs. The IABC method is a mixture of the original ABC algorithm and the DE algorithm, and allows the solution the opportunity to explore wider ranges. Moreover, the reward-based roulette wheel selection method was developed to replace the traditional roulette wheel selection method in IABC, which can strengthen the performance of the original ABC algorithm’s choice solution. Abstract—The artificial bee colony (ABC) algorithm is inspired by the behavior of honey bees. It is a relatively new optimization algorithm that has been proved competitive with conventional biology-inspired algorithms. The IABC algorithm is used, with the differential evolution (DE) algorithm added to the new solution search equation of ABC, to improve convergence speed. The IABC adopts the reward-based roulette wheel selection mechanism initially to divide all solutions suitably into feasible and infeasible solutions; thereafter, it divides them based on feasible and infeasible solutions for the implementation of incentives and punishments. Finally, the proposed method is applied to nonlinear system control problems. The experimental results of this study demonstrate the performance of IABC against that of other algorithms in nonlinear problems. II. I. INTRODUCTION x2 Neural fuzzy networks (NFNs) are powerful techniques and have been used to solve engineering problems [1]-[3] in recent decades. For the .
