摘要
针对粒子群算法在解决非线性类的误差评定中遇到收敛慢,易陷入局部最优的问题,且一般所提出的形状误差计算方法均针对特定模型的误差计算,通过对速度和搜索范围的相对调整,提出一种基于可变搜索区域的自适应粒子群优化算法.并将其用于球体、圆柱、圆锥的误差评定.对三类误差评定模型算法框架进行了统一描述,并给出各改进策略的实现要点.对比实验的计算结果表明,该方法对上述三类误差评定问题均有较好效果,在解决类似非线性优化问题时,其计算精度优于其他同类算法.
PSO algorithms usually encounter slow convergence and local optimum when they are used to solve nonlinear error evaluation problems. Target shapes of most exiting error calculation methods are just a certain specific model. A variable searching area and adaptive learning particle swarm optimization algorithm for sphere,cylinder,and cone error evaluation is proposed in this article. A unified framework for those three error evaluation models is defined and key points of the improvement strategy are described in detail. Experimental results show that the proposed error evaluation method has better results in solving similar nonlinear optimization problems compared to other similar algorithms.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第7期1615-1619,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(61103170)资助
福建省教育厅A类科技项目(JA12005)资助
关键词
粒子群优化算法
形状误差
圆锥度
圆柱度
球度
particle swarm optimization
shape error
conicity
cylindrical
sphericity
