| 应用BP网络校正的卡尔曼滤波器 |
| |
| 引用本文： | 朱志宇,张冰,姜长生.应用BP网络校正的卡尔曼滤波器[J].航天控制,2005,23(6):22-26. |
| |
| 作者姓名： | 朱志宇 张冰 姜长生 |
| |
| 作者单位： | 江苏科技大学电子信息学院,镇江,212003;南京航空航天大学自动化学院,南京,210016;南京航空航天大学自动化学院,南京,210016 |
| |
| 基金项目： | 船舶工业国防预研基金资助项目(编号为01J3.17) |
| |
| 摘 要： | 当目标作机动飞行时,由于未知机动引起的估计误差的影响,卡尔曼滤波器的跟踪性能将会严重下降。提出了一种应用BP神经网络对卡尔曼滤波结果进行校正的机动目标跟踪方法,BP网络可以辨识非线性机动,修正卡尔曼估计值,改善跟踪效果。同时还研究了BP网络的自适应算法。对两个变加速机动目标的跟踪结果表明,经过BP网络校正的卡尔曼滤波器不会发生误相关,跟踪的精度较高。
|
| 关 键 词： | 机动目标 跟踪 卡尔曼滤波器 BP神经网络 |
| 文章编号： | 1006-3242(2005)06-0022-05 |
| 修稿时间： | 2005年3月14日 |
Corrected Kalman Filter Using BP Neural Network |
| |
| Zhu Zhiyu,Zhang Bing,Jiang Changsheng.Corrected Kalman Filter Using BP Neural Network[J].Aerospace Control,2005,23(6):22-26. |
| |
| Authors: | Zhu Zhiyu Zhang Bing Jiang Changsheng |
| |
| Abstract: | During the process of tracking maneuvering target,the performance of Kalman filter will degrade because of unknown maneuver.One maneuvering target tracking method in which BP neural networks are applied to correct Kalman filter result is proposed in this paper.The function of BP networks is to identify nonlinear maneuvering,and correct Kalman estimated value so as to improve tacking results.Also the BP learning algorithm is studied as well.The racking results of two targets with varied acceleration show that wrong correlation do not happen to Kalman filter corrected by BP networks,and the tracking precision is very good. |
| |
| Keywords: | Maneuvering target Tracking Kalman filter BP neural network |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
