| 野值存在下的BP网络自适应卡尔曼滤波 |
| |
| 引用本文： | 李俊,周良荣,赵涛,王小峰.野值存在下的BP网络自适应卡尔曼滤波[J].火力与指挥控制,2009,34(11). |
| |
| 作者姓名： | 李俊 周良荣 赵涛 王小峰 |
| |
| 作者单位： | 西北工业大学航海学院,陕西,西安,710072 |
| |
| 基金项目： | 教育部新世纪人才支持基金资助项目 |
| |
| 摘 要： | 提出了一种抗野值的BP网自适应卡尔曼滤波算法.通过BP网络对新息序列的方差和均值的速率进行在线实时监测计算,输出一组加权系数对新息做修正处理,从而有效地抑制了野值对滤波器的影响.仿真证明该算法保持了新息的统计特性,提高了滤波器的精度和稳定性.
|
| 关 键 词： | BP神经网络 卡尔曼滤波 新息 残差 野值 |
BP Neural Network Adaptive Kalman Filtering When Existing Outliers |
| |
| LI Jun,ZHOU Liang-rong,ZHAO Tao,WANG Xiao-feng.BP Neural Network Adaptive Kalman Filtering When Existing Outliers[J].Fire Control & Command Control,2009,34(11). |
| |
| Authors: | LI Jun ZHOU Liang-rong ZHAO Tao WANG Xiao-feng |
| |
| Abstract: | An arithmetic of BP Neural Network adaptive kalman filtering is proposed for restraining the effect of outliers According to supervise and compute the velocity of means and covariance of innovation sequence by BP Neural Network online, a group of weighting coefficient is exported to modify the innovations Accordingly, the innovations' effect is restrained effectively The simulation results prove that the arithmetic can keep the statistic feature of innovation and increase the precision and stability of the filter. |
| |
| Keywords: | BP neural network Kalman filtering innovation residual difference outliers |
| 本文献已被 万方数据 等数据库收录！ |
|
