An adaptive parameter tuning of particle swarm optimization algorithm
An adaptive parameter tuning of particle swarm optimization algorithm
复制标题
粒子群优化算法的自适应参数整定
DOI：
10.1016/j.amc.2012.10.067
复制
发表时间：
2013-01-01
影响因子：
4
通讯作者：
Xu, Gang
中科院分区：
文献类型：
--
作者：
Xu, Gang
提出了一种基于速度信息的粒子群优化算法（APSO-VI）。本文首先分析了粒子的速度收敛性，指出了粒子速度与搜索失败率之间的关系，揭示了粒子群算法全局搜索能力较差的原因。然后，该算法引入了速度信息，定义为所有粒子速度的平均绝对值。提出了一种新的控制策略，即通过反馈控制，根据速度的平均绝对值跟随给定的非线性理想速度，动态地调整惯性权重，避免了速度在初始阶段趋近于零。在非线性理想速度的指导下，APSO-VI能够保持适当的种群多样性，有效地缓解早熟收敛。在一些基准函数上进行了数值实验，比较了该算法与不同的PSO变体。实验结果表明，该算法显著提高了粒子群算法跳出局部最优解的能力，显著提高了收敛速度和精度。(C)2012 Elsevier Inc. All rights reserved.
An adaptive parameter tuning of particle swarm optimization based on velocity information (APSO-VI) algorithm is proposed. In this paper the velocity convergence of particles is first analyzed and the relationship between the velocity of particle and the search failures is pointed out, which reveals the reasons why PSO has relative poor global searching ability. Then this algorithm introduces the velocity information which is defined as the average absolute value of velocity of all the particles. A new strategy is presented that the inertia weight is dynamically adjusted according to average absolute value of velocity which follows a given nonlinear ideal velocity by feedback control, which can avoid the velocity closed to zero at the early stage. Under the guide of the nonlinear ideal velocity, APSO-VI can maintain appropriate swarm diversity and alleviate the premature convergence validly. Numerical experiments are conducted to compare the proposed algorithm with different variants of PSO on some benchmark functions. Experimental results show that the proposed algorithm remarkably improves the ability of PSO to jump out of the local optima and significantly enhance the convergence speed and precision. (C) 2012 Elsevier Inc. All rights reserved.