文法演化是遗传程序设计的重要变体，易于描述和引入领域知识，被认为是解决金融预测、多目标决策、软件工程领域许多复杂问题的有力工具之一。作为计算密集型建模工具，它所面临的公开性难题在于缺乏科学的个体表示和系统化的评估方法。项目旨在模型方法基础上解决表示的语义复合性问题，进而提供利于多种分析方法引入的系统化解决方案。我们的研究任务主要包括新型演化架构、模块描述方法、语义演算、高性能并行计算技术以及其它相关问题。工作表明，模型方法的好处在于表达力强，能有效揭示形式概念、解的推导和演化的不确定性之间的关系，从而能为高性能文法演化计算系统的实现开辟新的途径。

Grammatical evolution (GE), an important variant of genetic programming (GP) which has the advantages of easy delineation and use of domain knowledge, is one of the major tools for coping with complex problems in the areas of financial prediction, multi-objective decision making, software engineering, etc. As a computation-intensive modeling tool, its major challenging issues with evolutions come mainly from lacking scientific representation and systematic evaluation methods. The purpose of the project is to establish a modeling framework from which systematic problem solving approaches using both compositional semantics and multiple analytical methods are easy to be achieved. Our major concerns in the present project cover the novel evolutionary framework, module describing, a calculus of semantics, high performance parallel computing technologies, and other related issues. The work reveals that model approaches have the strong expressiveness for giving a vivid view of the relationship among grammatical concepts, solution derivations, and uncertainties of evolutions, thus having the potential to provide new ways for implementing high performance computing systems of grammatical evolution.