| 基于Elman神经网络的多目标跟踪技术的研究 |
| |
| 引用本文： | 商庆华,邹海英,曹硕,吴丽华.基于Elman神经网络的多目标跟踪技术的研究[J].信息技术,2008,32(9). |
| |
| 作者姓名： | 商庆华 邹海英 曹硕 吴丽华 |
| |
| 作者单位： | 1. 哈尔滨理工大学测控技术与通信工程学院,哈尔滨,150040
2. 中国电子科技集团第二十二研究所,新乡,453000 |
| |
| 基金项目： | 黑龙江省教育厅科学技术研究项目 |
| |
| 摘 要： | 首先叙述了Elman神经网络的结构、原理和学习方法.针对Elman网络的学习率对网络收敛速度及稳定性影响很大,提出了一种可以自适应调整学习速率的改进的Elman网络学习算法;并基于Elman神经网络,采用模糊推理进行数据关联的方法,结合扩展卡尔曼滤波,提出一种新的多目标跟踪方法.最后应用此方法采用两个传感器对两个运动目标进行跟踪实验,并与BP神经网络对比得出仿真结果,实验结果表明,所提出的方法是一种可行的多目标跟踪方法.
|
| 关 键 词： | 多目标跟踪 Elman神经网络 模糊推理 扩展卡尔曼滤波 |
Study of multitarget tracking based on Elman neural network |
| |
| SHANG Qing-hua,ZOU Hai-ying,CAO Shuo,WU Li-hua.Study of multitarget tracking based on Elman neural network[J].Information Technology,2008,32(9). |
| |
| Authors: | SHANG Qing-hua ZOU Hai-ying CAO Shuo WU Li-hua |
| |
| Affiliation: | SHANG Qing-hua1,ZOU Hai-ying1,CAO Shuo2,WU Li-hua1(1.College of Measure-control Technology & Communication Engineering,Harbin Universityof Science , Technology,Harbin 150040,China,2.The 22th Research Institute of China Electronics Technology Group Corporation,Xinxiang 453000,China) |
| |
| Abstract: | The structure,theory and the learning algorithm of Elman Neural Network has been introduced firstly.Since the study rate has a great impact on the convergence rate and the stability in Elman Neural Network,we propose an improved learning algorithm,which can modify the study rate adaptively.Furthermore,a new multi-target tracking method is brought out based on Elman Neural Network,making use of fuzziness theory to carry out data association and combining with Extend Kalman Filter(EKF).At last,according to th... |
| |
| Keywords: | multi-target tracking Elman neural networks fuzziness theory extended Kalman filter |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
