Modified Brain Storm Optimization Algorithms Based on Topology Structures
An algorithm performs better often due to its communication mechanisms. Different types of topology structures denote various information exchange mechanisms. This paper incorporates topology structure concept into brain storm optimization (BSO) algorithm
- PDF / 493,967 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 61 Downloads / 300 Views
2
College of Management, Shenzhen University, Shenzhen 518060, China [email protected], [email protected] Department of Mechanism Engineering, Hong Kong Polytechnic University, Kowloon, Hong Kong
Abstract. An algorithm performs better often due to its communication mechanisms. Different types of topology structures denote various information exchange mechanisms. This paper incorporates topology structure concept into brain storm optimization (BSO) algorithm. Three types of topology structures, which are full connected, ring connected and star connected, are introduced. And three novel modified optimization algorithms based on topology structures are proposed (BSO-FC, BSO-RI, BSO-ST). Unimodal and multimodal criteria functions are employed to verify the effectiveness of the raised algorithms. In addition, both the original BSO algorithm and bacterial foraging optimization (BFO) algorithm are selected as contrastive algorithms to expose the optimization capacity of the proposed algorithms. Experimental results show that all of the modified algorithms have better performance than the original BSO algorithm, especially the BSO-ST algorithm. Keywords: Brain storm optimization Topology structures Population-based optimization Mutation operator Gaussian mutation
1 Introduction Optimization problem is an important branch of modern management. In many years, people are trying their best to find better ways to solve these problems. In recent years, optimization algorithms based on the population have been extensively investigated. In contrast to algorithms based on single-point, for example, hill-climbing algorithm, population-based optimization algorithms do its jobs by communicating and competing with each other. Nowadays, population-based optimization algorithms are generally classified as swarm intelligence algorithms. There are many swarm intelligence algorithms, such as particle swarm optimization (PSO) [1], bacterial foraging optimization (BFO) [2], artificial bee colony optimization (ABC) [3], ant colony optimization (ACO) [4], etc. But all of them are just inspired by simple animals or insects, such as ants, bees, birds, etc. As a new type of swarm intelligence, brain storm optimization (BSO) was first proposed by Shi in [5, 6]. BSO is motivated by the most intelligent organisms, the human being. After that, many scholars have conducted research on this algorithm because of its excellent performance. © Springer International Publishing Switzerland 2016 Y. Tan et al. (Eds.): ICSI 2016, Part II, LNCS 9713, pp. 408–415, 2016. DOI: 10.1007/978-3-319-41009-8_44
Modified Brain Storm Optimization Algorithms
409
Some people do research on the parameters of this algorithm. According to the dynamic range in the iteration, reference [7] proposed a dynamic step-size strategy, which was dynamically changed in each iteration. Reference [8] made investigation on the parameters in the BSO to see how they affected the performance of BSO. In order to reduce the computation time, reference [9] implemented only several iteration
Data Loading...
