传统的Monte Carlo法在进行装配公差分析时,为了达到必要的计算精度,需要大量的样本,对于日益复杂的机械产品而言,其越来越难以满足研发周期短、精度要求高、计算迭代量大的特点。针对这个问题,本文引入Sobol序列代替Monte Carlo法中的伪随机数序列,并将其融入到产品的公差分析模型。在理论分析的基础上,以某单向离合器为例,验证该方法的有效性及优越性。
In order to achieve the necessary calculation accuracy, the traditional Monte Carlo method requires a lot of samples in assembly tolerance analysis, as for the increasingly complex mechanical products, it is increasingly difficult to satisfy short development cycle,high precision, mass iterative calculation. For this problem, It introduces Sobol sequence to substitutes the pseudo-random number sequence in Monte Carlo method and turn it into the product tolerance analysis model. Based on theoretical analysis, It takes a one-way clutch as an example to verify its effectiveness and superiority.