Abstract: In order to solve the shortcomings that particle swarm optimization algorithm is easy to fall into local optimum and form early-maturing, this paper proposes a new Dual Adaptive PSO algorithm（DAPSO）which based on the theory of complex adaptive system by introducing the concept of chaos and adaptivity. Firstly, it uses the Logistic equation to create chaotic sequence in the beginning of initializing population. Secondly, it uses nonlinear dynamic adjustment strategy to adjust the particle’s individual learning factor and social learning factor. Thirdly, it uses （0, 1） random uniform distribution to instead of decreasing inertia weight to adjust inertia weight w. Finally, it uses six high-dimensional single mode and multi-modal Benchmark test function to do a simulation and it makes a comparison with PSO, 2PSO and KPSO. The result shows that DAPSO algorithm is more effective than the original particle swarm optimization algorithm in solving the global optimal and it has a better performance on the accuracy and the efficiency than other algorithms.

Key words: theory of complex adaptive system, Dual Adaptive Particle Swarm Optimization（DAPSO） algorithm, Logistic equation, nonlinear dynamic adjustment strategy, （0, 1） random uniform distribution

摘要： 针对粒子群优化（PSO）算法易陷入局部最优，发生早熟这一问题，借鉴复杂适应系统（CAS）理论，将混沌和自适应引入到基本PSO中，形成一种双重自适应PSO算法（DAPSO）。该算法在初始化种群时，采用Logisitic方程产生混沌序列；在迭代过程中，通过非线性动态调整策略调整粒子个体学习因子和社会学习因子的大小，采用（0，1）随机均匀分布代替惯性权重递减的方法对[w]进行自适应取值来更新粒子的速度和位移，最终实现算法求解全局最优的目标。最后运用六个高维单模态和多模态Benchmark测试函数对该算法进行仿真，并与PSO，2PSO，KPSO算法进行对比。对比结果表明，该算法在求解全局最优解时，效果明显优于其他粒子群算法，在精确性和寻优效率上较其他算法表现尤为突出。

关键词: 复杂适应系统（CAS）理论;双重自适应粒子群优化（DAPSO）算法;Logisitic方程;非线性动态调整策略;（0, 1）随机均匀分布