摘要
环境建模技术是机器人自主导航研究中的一个关键问题。针对机器人自主环境建模过程中噪声信息处理问题,提出一种基于小波变换多尺度分析的改进阈值噪声滤除算法,并比较不同小波函数的去噪效果,实现准确环境建模。通过分析传感器信号的特性,利用小波变换的多尺度分析特性,运用改进阈值算法来抑制包含有噪声的传感器信息在不同尺度上的噪声小波系数,实现在重构信号中去噪的目的。通过Pioneer3-DX机器人平台验证了所提出的方法能有效的滤除噪声,提高环境建模的准确性。
Environment modeling is one of the most important issues to tackle in autonomous navigation. Dealing with signal de-nois-ing problems in the environment modeling of autonomous mobile robot,a signal de-noising method based on multi-scale analysis of wavelet transform by improved threshold function was proposed,and to compare the difference of the effects by different wavelet types. It realized the accurate environment modeling. Analyzed the character of sensor signal and using a character of multiscale analysis of wavelet transform analysis,the aim of reconstructed signal de-noising was realized by using a method that wavelet trans-form coefficients of noise-sensor signal were filtered by changing threshold on the different scale.Experiment result on Pioneer3-DX proves that the proposed method filters the noise-sensor signal efficiently,and improves the environment modeling precision efficiently.
出处
《微计算机信息》
2009年第34期85-87,共3页
Control & Automation
基金
国防科工委基础项目
基金申请人:李磊民
(项目名称
基金编号不公开)
关键词
噪声滤除
传感器信号
环境建模
多尺度分析
小波变换
de-noising
sensor signal
environment modeling
multi-scale analysis
wavelet transform
