基于信息博弈的多源信息融合方法

Method of multi-source information fusion based on information game

针对多传感器信息融合系统中的信息冲突,结合博弈论的原理,在考虑信息冲突环境下局中人最优策略和最优支付的情况下,提出了求解信息冲突环境的信息博弈融合方法.首先通过信息熵的形式将多信息器表示成多源信息,并将多源信息的信息冲突过程抽象成为信息博弈过程,多传感器系统通过信息博弈的过程实现整个系统的信息收益最大化.通过对空调机调速器运行工况评价的实验分析,证实该方法在保持调节性能的同时,能从冗余的多源信息角度对系统的被控参数做出调整,各种动态品质特性优于传统的方法.

Aiming at the conflict circumstance in multi-sensor information system, considering the optimized algorithm and the optimization payment in information conflict situation, and combining with the information game method, an information game fusion method is presented. Expressing information quantity of multi-sensor with information entropy, the conflict course of multi-source information is abstracted to information game, in order to realize information entropy maximum of multi-sensor system for the whole system. By analyzing the operation of velometer in the air conditioning speed governing system, it is shown that the method has a better turning performance and can tune all controlled parameters from the redundancy view. The dynamic characteristics are better than the traditional way.