采用投影螺旋搜索的改进粒子群算法
An Improved Particle Swarm Optimization Algorithm with Projective Spiral Searches

针对传统粒子群算法在求解高维决策空间问题时容易陷入局部最优的缺点,提出采用投影螺旋搜索的改进粒子群算法。该算法提出了一种基于投影空间的螺旋搜索粒子更新方式,并应用于粒子群算法中以解决早熟问题;为了增强寻优能力,引入混沌策略生成螺旋搜索的参数以提高搜索的随机性;使用自适应算子选择策略分配螺旋搜索更新和传统粒子更新的选择概率,使2种粒子更新方式在不同搜索阶段发挥最大效用。仿真实验表明:与基本粒子群算法相比,提出的算法能够以较少的迭代次数收敛,寻优精度最大可提高10-13,适合于求解一类具有连续空间复杂多峰值特点的工程应用问题。
An improved PSO algorithm with spiral search(SSPSO) is presented to solve the shortcoming that the traditional particle swarm optimization (PSO) algorithm is easy to fall into a local optimum when complex functions in high？？dimensional decision space is solved. A new particle update method based on spiral searches in projection space is proposed and applied to PSO to solve the premature convergence problem; Chaos perturbation is introduced to improve the randomness of spiral search so that the search ability is improved. An adaptive selection strategy of operators is also used to balance the usage of spiral search and basic particle update strategy and to maximize their effectiveness in different search stages. Experimental results and a comparison with the traditional PSO indicate that SSPSO converges with fewer iterations and improves the computation accuracy up to 10-13. The algorithm is suitable for solving engineering application problems with characteristics of complexity, multiple peaks and continuous space