病态场景下多传感器系统误差的岭估计方法

发布时间：2018-04-10 15:32

本文选题：系统误差估计 + 岭估计　；参考：《系统工程与电子技术》2017年12期

【摘要】：多传感器系统误差估计是数据融合系统获得性能优势的关键前提之一。针对病态场景下传统系统误差估计方法数值不稳定的问题,对目标密集型和传感器密集型两种典型病态场景进行了理论分析,提出了多传感器系统误差的岭估计方法,以牺牲估计器无偏性的代价来改善估计结果的数值稳定性。通过引入条件数约束,给出了岭参数的最优取值方法。仿真结果表明,所提岭估计器在良态场景下与传统最小二乘估计器性能保持一致;在目标密集型场景下,与传统方法相比有显著性能优势;在传感器密集型场景下,对距离系统误差的估计性能有明显改善。
[Abstract]:Multi-sensor system error estimation is one of the key prerequisites for data fusion systems to gain performance advantages.In order to solve the problem of numerical instability of traditional system error estimation methods in ill-conditioned scenario, two typical pathological scenarios, target intensive and sensor intensive, are theoretically analyzed, and a ridge estimation method for multi-sensor system error is proposed.The numerical stability of the estimator is improved at the expense of unbiased estimator.By introducing the constraint of conditional number, the optimal value method of ridge parameter is given.The simulation results show that the proposed ridge estimator is consistent with the traditional least squares estimator in the good scenario, the performance of the proposed estimator is significantly superior to the traditional method in the target intensive scenario, and the performance of the proposed estimator in the sensor-intensive scenario is similar to that of the traditional least squares estimator.The estimation performance of distance system error is improved obviously.
【作者单位】：海军工程大学电子工程学院;中国人民解放军91715部队;
【基金】：第61批中国博士后科学基金面上资助(2017M613370)资助课题
【分类号】：TP212

【相似文献】