两类群体智能算法及其应用研究-计算机软件与理论专业论文.docxVIP

Ab Abstract PAGE PAGE III Abstract Swarm intelligence is defined as an attempt to design algor ithms and distributed problem-solving devices inspired by collect ive behavior of the socia l insert colonies and other anima l societ ies. Swarm intelligence is distributed, robust, stimer gy and simple. With these character ist ics, swarm intelligence show great advantages and potentia l in some research doma ins, such as solving the problem of optimizing combination, know ledge discovery, telecommunication network and robotics. Ant algor ithm and particle swarm optimization (PSO) algor ithm are two important content of swarm intelligence algor ithms. They have applied into more fields as they were deeply researched. Particle swarm optimization algor ithm is an evolutionary computation technique based on swarm intelligence optimization algor ithm, which was developed by Kennedy and Eberhart in 1995 and inspired by the socia l behavior of birds block ing. Due to its strong ability to globa l search, less parameters and simplicity , it has gained much attention since it is proposed. And it is develope d rapid ly and is used in ma ny fields such as funct ion optimization, neura l network training, fuzzy system control and other applied fields of genet ic algor ithm. But the basic particle swarm optimization (bPSO) has some demer its, such as relapsing into loca l extremum, slow convergence velocity and low convergence precision in the late evolutionary. The simple PSO(sPSO) discards the particle velocity and improves extraordinarily the convergence velocity and precision in the evolutionary optimization .and it looks more legible and terse. A modification to dynamica lly Decreasing inertia weight strategy in this article .It is demonst rated that there are evident super iorities in computationa l precision ,searching speed and steady convergence. Ant algor ithm is anot her important approach of swarm intelligence algor ithms. After the algor ithm being proposed,