为了提高光电跟踪系统的稳定性和自主跟踪能力，对多传感器数据融合算法和数据的有效性估计进行研究，提出一种多传感器自主跟踪算法.首先按照统计学方法，实时估计各传感器数据的误差协方差.然后按照均方误差最小准则，对各路数据进行融合.将最小二乘多项式拟合法和记忆衰减因子应用到误差协方差估计中，提高了融合结果的可信度.最后，提出一种多传感器跟踪数据切换策略，自动选择有效传感器数据中置信度最高、跟踪效果最优的一路数据，从而实现自主稳定跟踪的目的.实验结果表明，使用改进后的数据融合算法比原始方法的最终传感器选择结果正确率提高37．5％左右.在几种典型的传感器数据异常情况下，该数据融合算法和多路数据切换策略能够完成自主跟踪的目的.
In order to improve the stability and the ability of autonomous tracking of electro-optical tracking system,multi-sensor data fusion algorithm and estimate of the availability of sensor data is researched.In this paper a multi-sensor data fusion algorithm for autonomous tracking was put for-ward.First,the variance was estimated according to the statistics data real timely.Secondly,the fu-sion of the sensor data was made based on the rule of least mean square error.Then an improved esti-mation formula of the variance based on least square polynomial fit and memory attenuation factor was put forward which can improve the confidence interval of the fusion result.Finally,a multi-sensor switch-over strategy was put forward which can choose the sensor data that has the highest confidence interval and optimal tracking result from the valid ones autonomously.By this way we can achieve the aim of autonomous and stable tracking.Experiments results prove that the accuracy of sensor choice
using the improved formula was improved 37.5% than the previous algorithm.This algorithm of data fusion and choice strategy can realize autonomous track in several typical scene of sensor-invalid.
参考文献
| [1] | Julier S.;Uhlmann J.;Durrant-Whyte H.F..A new method for the nonlinear transformation of means andcovariances in filters and estimators[J].IEEE Transactions on Automatic Control,20003(3):477-482. |
| [2] | M. Sanjeev Arulampalam;Simon Maskell;Neil Gordon;Tim Clapp.A Tutorial on Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking[J].IEEE Transactions on Signal Processing: A publication of the IEEE Signal Processing Society,20022(2):174-188. |
| [3] | Djuric P.M.;Kotecha J.H.;Jianqui Zhang;Yufei Huang;Ghirmai T.;Bugallo M.F.;Miguez J..Particle filtering[J].IEEE Signal Processing Magazine,20035(5):19-38. |
| [4] | 高青 .多传感器数据融合算法研究[D].西安电子科技大学,2008. |
| [5] | 何友;陆大(纟金);彭应宁;熊伟.带反馈信息的分布多传感器航迹融合算法[J].电子科学学刊,2000(5):705-714. |
| [6] | 程国英;张淑梅;蔡盛;高慧斌;乔彦峰.实时特征级数据融合系统在光电经纬仪上的应用[J].光学精密工程,2008(11):2151-2157. |
| [7] | 苏秀琴;郝伟;李哲.一种基于光电经纬仪的数据预测跟踪技术[J].光子学报,2008(7):1464-1467. |
| [8] | 杜超;刘伟宁;刘恋.一种基于卡尔曼滤波及粒子滤波的目标跟踪算法[J].液晶与显示,2011(3):384-389. |
| [9] | 张品;董为浩;高大冬.一种优化的贝叶斯估计多传感器数据融合方法[J].传感技术学报,2014(5):643-648. |
| [10] | 龙辉平;习胜丰;侯新华.实验数据的最小二乘拟合算法与分析[J].计算技术与自动化,2008(3):20-23. |
| [11] | 王威立;郭劲;曹立华;陈娟.基于神经网络极限学习机数据融合的共轴跟踪[J].光学精密工程,2013(3):751-758. |
| [12] | 陈大庆;韩九强;于志坚.融合距离信息的红外自适应相关跟踪[J].光学精密工程,2010(8):1862-1868. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%