针对概率分布参数中存在区间的混合不确定问题,构造一种含区间变量的时变可靠性分析方法,可处理某些结构参数因样本不足无法获得精确概率分布情况下的可靠性计算问题。该方法首先将含区间变量的时变功能函数随机过程进行离散,并将之转换为一概率-区间混合静态可靠性问题。为求解混合可靠性计算中的双层嵌套优化,采用一种高效序列迭代格式,最终获得时变可靠度的上、下边界。最后以某机械零件、机械扭杆和齿轮传动为例,分别得出抗力和载荷均随时间变化且其概率分布参数中存在单个和多个区间不确定性参数情况下的时变可靠度曲线,验证了该方法的有效性。
The probability-interval hybrid reliability analysis method for time-variant structures is proposed to deal with time-variant reliability analysis of hybrid structures with random and interval variables, which can be used to solve the reliability without enough samples to construct precise random distributions. In this method, the stochastic process of the limit-state function with interval parameters firstly is discretized, and then simplified into the problem of time-independent reliability analysis for structures with probability-interval hybrid uncertainty. An efficient scheme sequential iteration also is used to solve the two-layer nesting optimization problem, and the upper and lower bounds of time-dependent reliability are obtain. A mechanical component, a torsion-bar and a gear reliability problems with time-dependent resistance and loads, as well as single or multiple interval distribution parameters, are investigated as examples, and the corresponding curves of reliability and failure probability bounds are obtained. The results show the effectiveness of the present method.