基于CAS理论的改进PSO算法

发布时间：2018-05-20 20:39

本文选题：复杂适应系统(CAS)理论 + 双重自适应粒子群优化(DAPSO)算法　；参考：《计算机工程与应用》2017年05期

【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。
[Abstract]:Aiming at the problem that particle swarm optimization (PSO) algorithm is prone to fall into local optimum and precocity occurs, chaotic and adaptive algorithms are introduced into the basic PSO to form a double adaptive PSO algorithm based on the theory of complex adaptive system (CAS). When initializing the population, the Logisitic equation is used to generate chaotic sequences, and in the iterative process, the particle individual learning factor and the social learning factor are adjusted by nonlinear dynamic adjustment strategy. The random uniform distribution is used to update the velocity and displacement of the particle instead of the decreasing inertia weight. Finally, the algorithm is used to solve the global optimal target. Finally, six high-dimensional single-mode and multi-modal Benchmark test functions are used to simulate the algorithm, and the results are compared with the PSO2PSO-KPSO algorithm. The results show that the proposed algorithm is more effective than other particle swarm optimization algorithms in solving the global optimal solution, especially in terms of accuracy and optimization efficiency.
【作者单位】：上海理工大学管理学院;上海理工大学超网络研究中心(中国);
【基金】：国家自然科学基金(No.71171135) 上海市一流学科建设项目(No.S1201YLXK) 上海市高原学科(管理科学与工程)建设项目上海高校青年教师培养资助计划项目(No.slg14020) 上海理工大学国家级项目培育课(No.15HJPY-QN09) 上海市哲学社会科学规划课题(No.2016EGL007)
【分类号】：TP18

【相似文献】