论文部分内容阅读
目的:分析由制造过程产生的变化对机构行为造成的影响。主要分析线性化方法对失败率估计的影响,从而平衡模型线性化造成的误差、计算时间和结果准确性。创新点:简化优化步骤,将机构行为线性化为几个部分,取代整体蒙特卡洛法,并采用一种迭代算法得到更加精确的结果。方法:1.采用带有非线性约束条件的几何线性化方法和基于失败率置信区间的算法(图3、4和5);2.以一个器件连接器为例,验证该算法在估计装配失败率上的作用。结论:1.线性化方法不影响蒙特卡洛仿真时间;2.线性化次数对计算时间和失败率估计准确率有很大影响(表1);3.线性化迭代统计方法相对于蒙特卡洛法在计算时间、计算精度和计算效率上有很大的优越性。
Purpose: To analyze the impact of changes resulting from the manufacturing process on organizational behavior. Mainly analyzes the influence of the linearization method on the estimation of the failure rate so as to balance the error caused by the linearization of the model, the calculation time and the accuracy of the result. Innovations: Simplify optimization steps by linearizing institutional behavior into several parts, replacing the overall Monte Carlo method and using an iterative algorithm to get more accurate results. Methods: 1. Using the geometric linearization method with nonlinear constraints and the algorithm based on the failure rate confidence interval (Figures 3, 4 and 5); 2. Taking a device connector as an example, Rate the role. Conclusion: 1. The linearization method does not affect the Monte Carlo simulation time; 2. The number of linearization has a great influence on the estimation accuracy of the calculation time and the failure rate (Table 1); 3. Linearization iterative statistical method compared with Monte Carlo The method has great advantages in calculating time, calculating accuracy and calculating efficiency.