Abstract: Gene Expression Programming（GEP） is the latest achievement in Evolutionary Algorithm（EA）, inheriting advantages such as simple encoding and powerful search capability from Genetic Algorithm（GA） and Genetic Programming（GP） respectively. This paper proposes a new GEP decoding method：GEP non-physical tree decoding algorithm, which can remarkably increase the decoding speed without affecting other aspect of GEP, and partially solve the bottleneck of GEP that huge time and space resource is consumed in its tree based evaluation processes.

Key words: Gene Expression Programming（GEP）, GEP decoding algorithm, non-physical tree GEP decoding algorithm, genetic algorithm, genetic programming

摘要： 基因表达式编程（Gene Expression Programming）是进化算法的最新成果。它继承了遗传算法（GA）编码简单与遗传程序设计（GP）有巨大空间搜索能力的优点。提出一种新的GEP解码方法：GEP的非物理树解码算法。其在不影响原算法其他性质的情况下极大地提高了传统解码算法的运行速度，在一定程度上解决了GEP进化过程中表达式树（Expression Tree，ET）建立和释放消耗巨大时空资源的瓶颈。

关键词: 基因表达式编程, GEP解码, 非物理树GEP解码, 遗传算法, 遗传程序设计