摘要
针对可靠性框图构造系统可靠度函数时需不交化处理的不足,提出了基于T-S故障树算法构造系统故障概率函数并结合可靠性费用函数构造可靠性优化模型的方法,降低了构造复杂系统可靠性优化模型的难度。针对微粒子群优化(μPSO)算法局部收敛性差、粒子群优化(PSO)算法全局搜索能力弱的不足,将μPSO算法和PSO算法进行综合,并结合死亡罚函数法构造了适应度函数,提出了混合μPSO算法,即μPSO-PSO算法。结合串联系统和桥式系统可靠性优化实例,考虑不同的粒子个数,证明了混合μPSO算法的优化结果比PSO算法、μPSO算法及PSO-μPSO算法的优化结果更为理想。
To overcome the shortages of disjoint-processing in system reliability function constructed by reliability block diagram,and to reduce the difficulties of constructing reliability optimization model,a method of system fault probability function constructed by T-S fault tree algorithm was presented,then the reliability optimization model was constructed combining with the reliability cost
function.To solve the shortcomings of the local convergence in the μPSO algorithm and global search ability in the PSO algorithm,the μPSO algorithm and the PSO algorithm were integrated,and the fitness function was constructed combining with death penalty function method,therefore a hybrid μPSO algorithm (μPSO-PSO algorithm) was proposed.The optimization results of the hybrid μPSO algorithm are better than the PSO algorithm,the μPSO algorithm and the PSO-μPSO algorithm.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2013年第18期2415-2420,共6页
China Mechanical Engineering
基金
国家自然科学基金资助项目(50905154)
河北省自然科学基金资助项目(E2012203015)
高等学校博士学科点专项科研基金资助项目(20091333120005)
河北省教育厅资助科研项目(ZH2012062)
关键词
可靠性优化
T-S故障树
混合μPSO算法
故障概率函数
死亡罚函数
reliability optimization
T-S fault tree
hybrid μPSO algorithm
fault probability function
death penalty function