摘要
为分析机械优化设计中粒子群算法(particle swarm optimization, PSO)的应用问题,在分析PSO经典形式的基础上,研究自适应改变惩罚系数的改进粒子群算法。通过列举测试函数,借助Matlab编程得出函数的收敛曲线,分析其收敛速度和收敛过程,表明该算法具有原理简单、实现容易、精度高、收敛速度快等优点。在机械优化设计中,以汽车差速器为例,用改进的粒子群算法对差速器体积进行优化,进一步验证该算法的可操作性和实用性。
To apply particle swarm optimization(PSO) to mechanical optimization design, the paper transforms the classical algorithm to an improved one through the introduction of adaptive change of penalty coefficient, obtains function's convergence curve after test function enumeration and Matlab programming, and analyzes its convergence speed and process,showing that the improved algorithm has the advantages of simple principle, high precision, fast convergence and easy implementation. With the automotive differential mechanism as an example, the optimization of its volume further validates that improved algorithm's practicality.
引文
[1]刘衍民.粒子群算法的研究及应用[D].济南:山东师范大学,2011.
[2]王金全,黄丽,杨毅.基于多目标粒子群算法的微电网优化调度[J].电网与清洁能源,2014,30(1):49-54.
[3]王国强,赵凯军,崔国华.机械优化设计[M].北京:机械工业出版社,2009:111-114.
[4]张辉.基于改进粒子群算法的可靠性优化研究[D].长沙:湖南大学,2012.
[5]陈黎卿,何钦章,郑泉,等.基于遗传算法的差速器多目标优化设计[J].机械制造,2007(5):25-27.
[6]姜平.基于Matlab的圆锥齿轮差速器优化设计[J].中国重型装备,2007(3):33-34.