随着人机交互技术的发展，可视化语言得到了越来越广泛的研究和应用。作为一种二维的形式化方法，图文法为可视化语言提供了直观而规范的描述手段。然而，目前大多数图文法理论框架的空间语义机制在通用性及分析性能方面存在着缺陷，影响了图文法的表达能力及其应用范围。针对现存问题，本课题拟对图文法形式框架下的空间语义机制进行研究。首先，为空间语义定义统一的离散坐标理论，并基于离散坐标与传统连续坐标之间的层级关系提供可选择的多粒度描述方案；其次，在此基础上构建一个具有二维/三维空间语义处理能力的图文法形式框架，并尝试利用空间语义关系的约束缩小归约过程中图柄的搜索空间，减少或消除空间图归约过程中的回溯，从而降低语法与语义分析的复杂性；最后，还将对新框架在工业设计领域中的应用进行探索，提出一种支持双向工作流的逻辑性图模型设计方法。

With the development of human-computer interaction technology, Visual Programming Languages have got more and more extensive research and applications. As a two-dimensional formal method, Graph Grammar provides an intuitive and formal way to specify Visual Programming Languages. However, the spatial semantic mechanisms of most existing Graph Grammar Formalisms have some deficiencies in the generality and analysis capabilities, which influences the expressive power and practical application scope of Graph Grammar. For solving the problems, this project intends to study the spatial semantic mechanism under Graph Grammar Formalism. We will first, define an unified theory of discrete coordinates, providing alternative multi-granularity descriptions schemes for spatial semantics by the hierarchical relationship between discrete coordinates and continuous coordinates; second, based on that, propose a new Graph Grammar Formalism with the ability of processing two or three dimensional spatial semantics and try to use the spatial relationships to narrow down the search space of redex and reduce or eliminate the back-tracking in spatial graph reduction, thus reducing the complexity of syntax and semantics parsing; finally, explore the practical applications of the new formalism in industrial design and propose a bi-directional logical design method for graph models.