混沌粒子群算法及其在桁架结构优化设计中的应用

Chaos-based Particle Swarm Optimization Algorithm and its Application in Truss-optimization Design

文章提出一种基于混沌思想的粒子群优化算法,它利用粒子群优化算法收敛速度快和混沌运动遍历性、随机性等特点,对标准粒子群优化算法进行了改进。在算法的初始化阶段,对粒子的位置混沌初始化;在算法运行过程中,对粒子的位置进行混沌更新和粒子群更新相结合的更新方式,使全局收敛与局部收敛达到一定平衡。将该算法应用于桁架结构的优化设计,与标准粒子群算法优化的结果相比较,混沌粒子群算法提高了对多维空间全局搜索能力,可有效避免早熟收敛现象。

Chaos-based particle swarm optimization algorithm （PSOA）, which makes use of the advantages of PSOA such as rapid convergence, transverse chaos movement and randomness, is presented through the improvement of standard PSOA. At the initialization stage of algorithm, the location of the particle is initialized by chaos. During the running time, the particles are updated through the combination form of chaos update and particle swarm update so that the global convergence could balance against local convergence. Such algorithm is used in truss-optimization design. Compared with the results from standard PSOA, chaos-based PSOA has better global search capa- bility for the multi-dimensional variable space and can effectively avoid the premature convergence.