基于神经网络的多传感器融合技术研究  

The Study of Multi Sensor Fusion Technology Based on the Neural Network

作　　者：黎果[1] 李志蜀[1] 李奇[2] 李建宏[1] 

机构地区：[1]四川大学计算机学院,610064 [2]四川师范大学,610000

出　　处：《微计算机信息》2009年第16期145-147,共3页Control & Automation

摘　　要：研究了基于神经网络的多传感器融合技术,并将其应用于自主吸尘机器人中。给出了神经网络传感器融合技术的基本原理,探索了改进的BP信息融合算法,使得改进后的算法在收敛速度上得到了很大的改善;介绍了传感器的监测系统和传感器特征的选择,构建了BP算法模型,并进行了仿真,得到了良好的实验结果。This paper investigates a type of multi-sensor technology based on the Neural Network, and applies it in the autonomous cleaning robots. The basic principles of Neural Network Multi-sensor Fusion technology are afforded. In order to improve the rate of the convergence, the paper uses some improved measure in the BP Exploration information fusion algorithm. Then it gives out the sensor monitoring system and the choice of the sensor characteristics, constructs the model of the BP algorithm, and presents the good resuh in the simulation.

关 键 词：神经网络 多传感器融合 BP算法 自主吸尘机器人 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]