建议先弄本书看看，易丹辉或者高铁梅的都可以。

步骤都差不多的

VAR和vecm有内在联系

差一个约束条件:)

有协整约束的var模型就是vecm。

楼上说的都有道理，可还是让人很困惑！

VAR is Vector auto regression model which is suitable when there is no coingegration among regressors.
VECM is kind of VAR that includes additional information among the cointegrated variables which is the equilibrium error term.
So, the difference between VAR and VECM is the equilibrim error term.
I'm not able to type Mandarin now, however if you have additional questions, you can message me.
Wish you good luck