摘要： 针对G 语言特定的数据要求，提出一种新的内存分配与管理算法，在满足多种基本类型和数组及簇等复杂数据结构的分配与管理的基础上，实现数据的产生及生命周期管理、数据信息的完备性及独立性描述。在不牺牲运行效率的前提下，能够长期稳定地运行，有效避免内存碎片的产生。并且将此算法在自主开发的虚拟仪器开发平台LabScene 中完整实现，在G 语言的解析与运行模型中发挥关键作用。

关键词: G 语言；虚拟仪器；内存分配与管理；数据流；LabScene

Abstract: Aimed at the graphic language specific data requirements, a kind of new memory allocation and management algorithm is presented in this paper. Based on allocation and management for basic data type and complicated data structure as array and cluster, it realizes the data generation and life cycle management, completeness of data information and independent character. On the premise of not sacrificing operational efficiency, the algorithm can run firmly for a long time, and avoid producing the memory fragments effectively. The paper realizes this algorithm integrallty in virtual instruments developing platform LabScene that is developed independently. This algorithm plays a key role in G language interpreter and running model.

Key words: Graphic language; Virtual instruments; Memory allocation and management; Dataflow; LabScene