基于无抽取Haar算法的实时卡尔曼滤波方法

来源期刊:中南大学学报(自然科学版)2011年第12期

论文作者:谭平 蔡自兴

文章页码:3760 - 3764

关键词:卡尔曼滤波;多尺度;小波变换;Haar算法;去噪

Key words:Kalman filter; multi-scale; wavelet transform; Haar algorithm; denoise

摘    要:针对基于分段方式的多尺度卡尔曼滤波其计算量大、延时长的问题,提出基于无抽取Haar算法的实时卡尔曼滤波方法,该方法采用简单的加减、移位运算在t时刻完成多尺度变换,然后在各个尺度进行小波阈值去噪和卡尔曼滤波;为了验证该方法的有效性,在自主改装的智能车上对低精度加速度传感器进行实验。研究结果表明:通过小波重构完成信号处理,提高了算法实时性,并且有效减少重复运算;实时卡尔曼滤波方法有效提高了传感器的性能,在不能准确估计状态转移误差情况下,该方法的去噪性能优于单独的卡尔曼滤波去噪性能。

Abstract: To address the problem of large computation and long delays in existing multi-scale Kalman filter, a real time Kalman filter based on the non-decimated Haar algorithm was proposed. A simple addition, subtraction and shift operation was used to complete multi-scale transformation at time t, and the signal was reconstructed after de-noising by wavelets soft-threshold and Kalman filter on each scale. To verify the validity of this method, the experiments with low-precision acceleration sensor in self-modified intelligent vehicle were carried out. The results show that it runs real time and more efficiently by reducing the repeat computation. Its performance is improved when the error of status can not be estimated accurately, and the de-noising performance of this method is superior to that of a single Kalman filter.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号