在虚拟现实教学平台的设计与实现过程中，为了实现教学过程的自动控制，有必要对常用教学模式进行建模。本文在借鉴前人对教学模式建模的经验基础上，针对教学过程复杂多变的特点，采用分层模拟和逐步求精的思想，利用Petri网对教学模式进行建模，并对建立的模型进行了性质验证。文章第一部分介绍了Petri网的基本理论和各项性质（可达性、活性、安全性和公平性），以及逐步求精的建模思想。第二部分介绍了几种常用的教学模式，并从中总结抽取了基本的教学元素。第三部分，对常用教学模式进行建模（重点针对有意义接受学习教学模式），先建立主干网模型，随后利用抽取的基本教学元素，构建子网模型，替换主干网中相应节点。然后，用构建可达标识树的方法，分别对主干网和各子网的性质进行了验证，并证明了用子网替换主干网节点的等价性。第四部分，文章结合工作流的基本原理，介绍了此研究的应用方向，并介绍了一个以Petri网为理论基础的工作流引擎——JBPM。

In the process of designing Virtual Reality teaching platform, In order to achieve automatic control of the teaching process, it is necessary to model the common teaching mode. As teaching is a complex course, so it is hard to model it in a single map. So this paper will use the “Stepwise refine” mode to model common teaching mode based on Petri Net. This paper has achieved the following accomplishments: First, introduce the basic theory of Petri Net and the “Stepwise refine” mode. Second, analyze some of the common teaching mode, and abstract the basic element in teaching process. Third, model the teaching mode of meaningful accept learning, and verify the character of the model built, such as reachability, boundedness, liveness, fairness. Last, introduce the basic theory of workflow in which field to use, and introduce a workflow engine based on Petri Nets—JBPM.