کلونی زنبور عسل خود تطبیقی مصنوعی برای بهینه سازی عددی جهانی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|10494||2012||7 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : IERI Procedia, Volume 1, 2012, Pages 59–65
The ABC algorithm has been used in many practical cases and has demonstrated good convergence rate. It produces the new solution according to the stochastic variance process. In this process, the magnitudes of the perturbation are important since it can affect the new solution. In this paper, we propose a self adaptive artificial bee colony, called self adaptive ABC, for the global numerical optimization. A new self adaptive perturbation is introduced in the basic ABC algorithm, in order to improve the convergence rates. 23 benchmark functions are employed in verifying the performance of self adaptive ABC. Experimental results indicate our approach is effective and efficient. Compared with other algorithms, self adaptive ABC performs better than, or at least comparable to the basic ABC algorithm and other state-of-the-art approaches from literature when considering the quality of the solution obtained.
Optimization problems play an important role in both industrial application fields and the scientific research world. During the past decade, we have viewed different kinds of meta-heursitic algorithms advanced to handle optimization problems. Among them, Meta-heuristic based methods, such as simulated annealing (SA), genetic algorithm (GA), particle swarm optimization algorithm (PSO), artificial bee colony (ABC), and differential evolution [1-4], may be one of the most popular methods. Particularly, artificial bee colony algorithm  is a population-based heuristic evolutionary algorithm inspired by the intelligent foraging behavior of the honeybee swarm. B. Akay and D. Karaboga  proposed a modify versions of the artificial bee colony. The modified artificial bee colony applied for efficiently solving real parameter optimization problem. The modified ABC algorithm employs a control parameter that determines how many parameters to be modified for the production of a neighboring solution. A scaling factor that tunes the step size adaptively was introduced. However, this field of study is still in its early days, a large number of future researches are necessary in order to develop the new version artificial bee colony algorithm for optimization problems. Since ABC is a particular instance of EA, it is interesting to investigate how self adaptive can be applied to it. Until now, there is no paper to focus on self-adaptive in ABC has been reported. In our paper, the parameter control technique is based on the self adaptive of the magnitudes of the perturbation, associated with the evolutionary process. The main goal here produces a flexible ABC, in terms of control parameter. We propose a self adaptive ABC which the variant of control parameter is changed according to the iteration. The low value of control parameter allows the search to find the optimal search in small step. However, it makes the convergence slower. A high value of control parameter speed up the search but it reduces the exploitation capability of the perturbation process. Therefore, we use this method which can balance the exploration and the exploitation of ABC.
نتیجه گیری انگلیسی
In this paper, we propose a self adaptive artificial bee colony, called self adaptive ABC. The proposed self adaptive method is an attempt to determine the values of control parameter ij . Our self adaptive ABC algorithm has been implemented and test on benchmark optimization problems taken from literature. 23 benchmark functions chosen from literature are employed. The results show that the proposed self adaptive ABC algorithm clearly outperforms the basic ABC. Compared with some evolution algorithms from literature, we find our algorithm is superior to or at least highly competitive with these algorithms. In this paper, we only consider the global optimization. The algorithm can be extended to solve other problem such as constrained optimization problems.