机构地区: 遵义师范学院数学系
出 处: 《计算机工程与应用》 2011年第12期23-26,共4页
摘 要: 针对约束优化问题的求解,提出一种改进的粒子群算法(CMPSO)。在CMPSO算法中,为了增加种群多样性,提升种群跳出局部最优解的能力,引入种群多样性阈值,当种群多样性低于给定阈值时,对全局最优粒子位置和粒子自身最优位置进行多项式变异;并根据粒子违背约束条件的程度,提出一种新的粒子间比较准则来比较粒子间的优劣,该准则可以保留一部分性能较优的不可行解;为提升种群向全局最优解飞行的概率,采取一种广义学习策略。对经典测试函数的仿真结果表明,所提出的算法是一种可行的约束优化问题的求解方法。 An improved particle swarm optimizer is proposed for solving constrained optimization problems(CMPSO).In or- der to increase the diversity of the swarm and improve the ability to escape from local optima, the diversity threshold value (λα) is introduced.When the swarm diversity value is equal or lesser than λα, the polynomial mutation is invoked for the global best performing particle(Gbest) and the particle personal best performing particle(Pbest).A new comparison strategy is proposed based on the violation degree of each particle;which can keep some infeasible solutions that have the good performance.To improve probability of flying to the optimal solution,a comprehensive learning strategy is adopted.The experiments on benchmarks indicate that the proposed algorithm is a feasible algorithm for solving constrained optimization problems
领 域: [自动化与计算机技术] [自动化与计算机技术]