Int. Conf. on Communication, Circuits and Systems (ICCCAS), Chengdu, China, 2002 ©2002 IEEE 1170 Hybird Particle Swarm Optimizer with Mass Extinction Xiao-Feng Xie, Wen-Jun Zhang, Zhi-Lian Yang Institute of Microelectronics, Tsinghua University, Beijing 100084, P.R.China Email: xxf@dns.ime.tsinghua.edu.cn Abstract - A hybrid particle swarm optimizer with mass extinction, which has been suggested to be an important mechanism for evolutionary progress in biological world, is presented to enhance the capacity in reaching optimal solution. The testing results of three benchmark functions that typically used in evolutionary optimization research indicate this method improves the performance effectively. 1. Introduction The particle swarm optimization (PSO) algorithm is an evolutionary computation technique that originally introduced by Kennedy and Eberhart in 1995 [1, 2]. The underlying motivation for the development of PSO algorithm was social behavior of animals such as bird flocking, fish schooling, and swarm theory [3]. Work presented in [4, 5] describes the complex task of parameter selection in the PSO model. Several researchers have analyzed the performance of the PSO with different settings, e.g., neighborhood settings [6], cluster analysis [7]. It has been used for approaches that can be used across a wide range of applications, as well as for specific applications focused on a specific requirement [8]. Comparisons between PSO and the standard GA were done analytically [9] and also with regards to performance [10]. Angeline [10] points out that the PSO performs well in the early iterations, but has problems reaching a near optimal solution in several real-valued function optimization problems. Both Eberhart [9] and Angeline [10] conclude that hybrid models of the standard GA and the PSO, could lead to further advances. Paleontological findings have revealed that mass extinction has been a common phenomenon in evolution [11]. It has been suggested to be an important mechanism for evolutionary progress in biological world [12], since extinction allows the repopulation of niches and gives space for new adaptations. In the field of evolutionary algorithms, this idea has been the motivation for so-called (mass-) extinction models, which has been introduced recently [13-15]. It is therefore natural to ask if mass extinction can be exploited to increase the efficiency of PSO. Here we review the role of mass extinction in the fossil record and simulate this process in a hybrid particle swarm optimizer with mass extinction. Both standard and hybrid versions are compared on three numerical optimization problems typically used in evolutionary optimization research. The preliminary results suggest that mass extinction can enhance the performance. 2. Standard particle swarm optimization (SPSO) The fundament to the development of PSO is a hypothesis [16] that socia l sharing of information among conspeciates offers an evolutionary advantage. PSO is similar to the other evolutionary algorithms in that the system is initialized with a population of random solutions. However, each potential solution is also assigned a randomized velocity, and the potential solutions, call particles, corresponding to individuals. Each particle in PSO flies in the D-dimensional problem space with a velocity which is dynamically adjusted according to the flying experiences of its own and its colleagues. The location of the i th particle is represented as X i = (x i 1 , …, x id , …, x i D ), where x id ∈[l d , u d ], d ∈[1, D], l d , u d are the lower and upper bounds for the dth dimension, respectively. The best previous position (which giving the best fitness value) of the i th particle is recorded and represented as P i = (p i 1 ,…, p id , …, p i D ), which is also called pbest. The index of the best particle among all the particles in the population is represented by the symbol g . The