贝叶斯数据融合中“模糊”数据随机集模型的探讨

'Discussion Bayesian data fusion "fuzzy" data set model of random

基于随机集理论运动物体轨迹的检测、提取、识别和跟踪是比较前沿的研究课题之一;有限集统计理论的核心是多源、多目标微积分计算。随着GPS、GIS的快速发展,为GPS、GIS提出了更高的要求;怎样处理具有多源、多目标传感器"模糊"数据是当前工程技术人员急需解决的问题。将贝叶斯方法论进行推广,利用有限集统计理论的专家系统方法(如模糊理论、D-S证据理论),可以使得不完整特性的传感器数据得到处理。

Based on the motion detection, random set theory object recognition and tracking is a research subject of frontier, the core of finite set statistics theory is to calculate the multi-source, multi object calcu-lus. With the rapid development of GPS GIS put forward higher require ments for GPS GIS, How to deal with multi-source, multi target sensor＆quot;fuzzy＆quot;data is the current solution of engineering and technical person-nel in urgent need of the problem. The Bias methodology is extended, the expert system method using the fi-nite set statistics theory （such as fuzzytheory, D-S theory of evidence）, It can make Incomplete Characteristic’ s Sensor data to get a better deal.