带有色厚尾量测噪声的鲁棒高斯近似滤波器和平滑器

发布时间：2018-01-08 02:05

本文关键词：带有色厚尾量测噪声的鲁棒高斯近似滤波器和平滑器　出处：《自动化学报》2017年01期 　论文类型：期刊论文

【摘要】：为了解决带有色厚尾量测噪声的非线性状态估计问题,本文提出了新的鲁棒高斯近似(Gaussian approximate,GA)滤波器和平滑器.首先,基于状态扩展方法将量测差分后带一步延迟状态和白色厚尾量测噪声的非线性状态估计问题,转化成带厚尾量测噪声的标准非线性状态估计问题.其次,针对量测差分后模型中的噪声尺度矩阵和自由度(Degrees of freedom,DOF)参数未知问题,设计了新的高斯近似滤波器和平滑器,通过建立未知参数和待估计状态的共轭先验分布,并利用变分贝叶斯方法同时估计未知的状态、尺度矩阵、自由度参数.最后,利用目标跟踪仿真验证了本文提出的带有色厚尾量测噪声的鲁棒高斯近似滤波器和平滑器的有效性以及与现有方法相比的优越性.
[Abstract]:In order to solve the problem of nonlinear state estimation with colored thick tail measurement noise, a new robust Gao Si approximation Gaussian approximate is proposed in this paper. First, the nonlinear state estimation problem with one-step delay state and white thick tail measurement noise after the measurement difference is proposed based on the state expansion method. It is transformed into a standard nonlinear state estimation problem of measurement noise with thick tail. Secondly, aiming at the noise scale matrix and degrees of freedom in the measured difference model, degrees of freedom are proposed. A new Gao Si approximate filter and smoother is designed for unknown DOF parameters. The conjugate prior distributions of unknown parameters and states to be estimated are established and the unknown states are estimated simultaneously by variational Bayes method. Scale matrix, degree of freedom parameter. Finally. The effectiveness of the proposed robust Gao Si approximation filter and smoother with color thick tail measurement noise and the superiority compared with the existing methods are verified by target tracking simulation.
【作者单位】：哈尔滨工程大学自动化学院;哈尔滨工业大学电气工程及自动化学院;中国舰船研究设计中心;
【正文快照】： 引用格式黄玉龙,张勇刚,武哲民,李宁,王刚.带有色厚尾量测噪声的鲁棒高斯近似滤波器和平滑器.自动化学报,2017,43(1):114-131Robust Gaussian Approximate Filter and Smoother withColored Heavy Tailed Measurement NoiseHUANG Yu-Long1ZHANG Yong-Gang1WU Zhe-Min2LI Ning1

【相似文献】