证据理论应用于多传感器（知识源）的信息综合时，辨识框（假设或命题空间）初始定义为有穷个穷举的不相容命题的集合。本文指出常用辨识框可分为符号集合型、有序命题集合型和实数闭区间型。并将证据理论信息综合原则扩展到辨识框为实数闭区间问题进行分析． When the evidence theory is applied to the information synthesis of multi-sensor (knowledge source), the identification box (hypothesis or proposition space) is initially defined as a collection of incompletible propositions with a finite number of exhaustions. This paper points out that commonly used identification box can be divided into symbolic set, ordered set and real closed set. The theory of evidence theory is extended to the problem that the identification box is closed real.