机构地区: 西南交通大学机械工程学院
出 处: 《起重运输机械》 2004年第6期31-35,共5页
摘 要: 针对复杂系统可靠性试验非常少甚至没有做的情况 ,提出了基于单元信息进行可靠性综合的方法。该方法不需假设系统或单元产品的寿命服从某一分布 ,减少了因寿命分布选择不当所造成的可靠性和方差的误差。在获得单元可靠性的均值和方差的基础上 ,利用系统可以分解为单一的串联或并联关系 ,通过逐步综合获得复杂系统的可靠性均值和方差估计值。利用系统信息熵原理 ,将部件的试验数据折合为系统的试验数据 ,获得系统的子样数。由此提出了小子样下的系统可靠性置信区间估计新方法 ,该方法只假设系统可靠性估计服从正态或对数正态分布。新方法使用限制少 ,计算简单 。 Considering that few or even no reliability tests are conducted for complex systems,a reliability comprehensive estimation method is provided based on component test data This method can be used without assuming that lives of system or component follow time-to-failure distribution,decreasing reliability and variance errors due to incorrect life distribution selection.The complex system reliability mean and variance can be obtained based on component reliability information and the condition that the complex system can be decomposed into series or parallel systems.The information entropy theory is applied to transform component test data into system test data.Thus a new small sample based system reliability confidence interval estimation method is offered under the assumption that the system follows normal or log normal distribution.The new method is characteristic of less constraints and simplified calculation and is suitable for less data system reliability analysis.
关 键 词: 复杂系统 可靠性 置信区间 小子样 方差估计 信息熵 串并联系统
领 域: [自然科学总论]