多因子异方差分析
Multi-Way Heteroscedastic Analysis
-
摘要: 提出一种多因子异方差分析方法,在变异系数相同的条件下,详细讨论了单因子、双因子以及多因子异方差分析。该方法可以对多个正态母体是否具有某种不同的均值进行检验和估计。并且通过少量试验,对仿真和数字化设计结果进行检验,分析各个因素(因子)对仿真和数字化设计结果的影响,以确定哪些因素已被正确仿真,而哪些因素尚未被正确模拟,从而指导仿真软件的编制。此外,该方法还可以根据试验数据对仿真和数字化设计结果进行修正。Abstract: A method for multi-way heteroscedastic analysis is presented,which detailedly discusses the heteroscedastic analysis of one,two and multiple factors,in the condition of the same coefficient of variation in all states.It can test whether the multiple populations have a certain kind of different means,and estimate their parameters.With small sample,it can test the results of simulation or digital design and analyze the influence of each factor in the results.And then,it can point out which factors have been simulated correctly and which haven’t.Furthermore,this method can correct the results of simulation or digital design according to the experimental data.
-
Key words:
- aerospace propulsion system /
- variance analysis /
- factor analysis /
- small sample /
- test /
- simulation /
- digital design
点击查看大图
计量
- 文章访问数: 1356
- HTML浏览量: 1
- PDF量: 448
- 被引次数: 0