【摘要】 火箭推进剂的燃烧理论长期落后于应用研究，如何从已有的大量配方数据中直接获取有用的信息，已成为实验工作者更关心的问题。我们建立了一套基于模式识别分析和物理化学分析的软件包，包括数据库及服务于数据库的各种算法库，除了进行推进剂的燃速数据处理、能量特性分析外，还能够从推进剂配方数据中抽象出大量特征，并通过这些特征与推进剂性能的相关性分析筛选出影响性能的主要因素，进而建立这些因素与推进剂性能的数量关系，以辅助推进剂燃烧特性和机理的研究。更多还原

【Abstract】 The theoretic studies of rocket propellant combustion developed not as so far as its application studies. How to get messages from a mass of known prescription data seems more important to those experimental explorers. A software system has been built up based on pattern recognition technique and physicochemical analysis method. It includes a database and kinds of algorithms. It treats burning rate data and calculates energy properties for the propellant. It extracts features from those prescription data, analyzes the relationships between these features and combustion properties of propellant. From these messages, it points out features that seem more important to control those properties. Some qualitative functions and rules will be built to show these relationships and can be expected to guide the propellant studies about its combustion mechanism and combusti0n properties.更多还原