Optimal foraging algorithm for global optimization

作     者：Zhu, Guang-Yu Zhang, Wei-Bo 

作者机构：Fuzhou Univ Coll Mech Engn & Automat Fuzhou 35002 Fujian Peoples R China 

出 版 物：《APPLIED SOFT COMPUTING》 (应用软计算)

年 卷 期：2017年第51卷

页      面：294-313页

核心收录：

学科分类：08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：natural science foundation of Fujian province [2014J01183] program of department of science and technology of Fujian province [2016H0015] 

主　　题：Optimal foraging algorithm (OFA) Optimal foraging theory Stochastic search algorithm Evolutionary algorithms Behavioral ecologya 

摘      要：An optimization algrothim,inspried by animal Behavioral Ecology Theory-optimal Foraging Theory named the optimal Foraging Algorithm (OFA) has been developed. As a new stochastic search algorithm, OFA is used to solve the global optimization problems following the animal foraging behavior. During foraging, animals know how to find the best pitch with abundant prey;in establishing OFA, the basic operator of OFA was constructed following this foraging strategy. During foraging, an individual of the foraging swarms obtained more opportunities to capture prey through recruitment;in OFA the recruitment was adopted to ensure the algorithm has a higher chance to receive the optimal solution. Meanwhile, the precise model of prey choices proposed by Krebs et al. was modified and adopted to establish the optimal solution choosing strategy of OFA. The OFA was tested on the benchmark functions that present difficulties common to many global optimization problems. The performance comparisons among the OFA, realcoded genetic algorithms (RCGAs), Differential Evolution (DE), Particle Swarm Optimization (PSO) algorithm, Bees Algorithm (BA), Bacteria Foraging Optimization Algorithm (BFOA) and Shuffled Frog-leaping Algorithm (SFLA) are carried out through experiments. The parameter of OFA and the dimensions of the multi-functions are researched. The results obtained by experiments and Kruskal-Wallis test indicate that the performance of OFA is better than the other six algorithms in terms of the ability to converge to the optimal or the near-optimal solutions, and the performance of OFA is the second-best one from the view of the statistical analysis. (C) 2016 Elsevier B.V. All rights reserved.