摘要
为了对液压板料折弯机的可靠性进行评估,提出了基于幂律过程的可靠性评估方法,建立了基于幂律过程的故障数据模型。对故障数据进行趋势检验和更新过程检验,验证了幂律过程应用的前提条件成立。应用极大似然估计法和Fisher信息矩阵法给出了模型参数的点估计与区间估计,并给出了折弯机可靠性指标的点估计和基于Delta法的区间估计。AIC信息准则和BIC信息准则表明,幂律过程优于威布尔分布和广义更新过程(GRP),能够反映出折弯机的瞬时与累积可靠性特性。
An approach of assessing reliability for hydraulic sheet metal bending machines based on Power Law Process （PLP） was pro- posed, and a PLP function for failure times was built. Trend test and renewal test were performed on the failure data to assure the validity of application for PLP. Point and confidence interval estimators were given by maximum likelihood method and Fisher information matrix. Point estimates of reliability metrics were also derived with their confidence bounds offered through Delta method. Akaike information and Bayesian information criterion demonstrate that PLP outperforms Weibnll distribution method and General Repair Process and is capable to capture the instantaneous and cumulative reliability properties of bending machines.
出处
《锻压技术》
CAS
CSCD
北大核心
2014年第1期74-79,共6页
Forging & Stamping Technology
基金
国家自然科学基金资助项目(50835008
51175527)
国家科技重大专项"高档数控机床与基础制造装备"(2010ZX04014-015
2011ZX04014-011
2012ZX04011-031)
重庆市教委资助项目(KJ121602)
关键词
可靠性评估
折弯机
幂律过程
Delta法
reliability assessment
bending machine
power law process
Delta method
