时变系统全局最优加权观测融合白噪声反卷积估值器研究.pdfVIP

Proceedings of the 25th Chinese Control Conference 7-11 August, 2006, Harbin, Heilongji ang 变系统全局最优加权观测融合白噪声反卷积估值器 孙小君，王佳伟，邓自立 黑龙江大学自动化系, 哈尔滨 150080 E-mail: dzl@hlju.edu.cn 摘 要: 白噪声反卷积或输入白噪声估计问题在石油地震勘探中有重要的应用背景. 对带多传感器和带不相关 噪声的线性离散时变随机系统,应用Kalman滤波方法,基于加权最小二乘法,提出了全局最优加权观测融合白噪 声反卷积估值器,可统一处理 入白噪声融合滤波、平滑和预报问题.一个Bernoulli-Gaussian 入白噪声融合平 滑器的Monte Carlo仿真例子说明了其有效性. 关键词: 时变系统, 加权观测融合, 白噪声估值器, 反卷积, 全局最优性, Kalman滤波方法 Globally Optimal Weighted Measurement fusion white noise deconvolution estimator for time-varying systems SUN Xiao-j un, WANG Jia-wei, DENG Zi-li Department of Automation, Heilongj iang University, Harbin 150080, China E-mail: dzl@hlju.edu.cn Abstract: White noise deconvolution or input white noise estimation problem has important application background in oil seismic exploration. For the linear discrete time-varying stochastic systems with multisensor and uncorrelated noises, a globally optimal weighted measurement fusion white noise deconvolution smoother is presented based on the method of weighted least squares, using Kalman filtering method, which can handle the white noise fusion filtering, smoothing and prediction problems in a unified framework. A Monte Carlo simulation example for a Bernoulli-Gaussian input white noise fused smoother shows its effectiveness. Key Words: time-varying system, weighted measurement fusion, white noise estimator, deconvolution, global optimality, Kalman filtering method 观测融合,即用增加观测向量维数方法合并各传感 1 引言(Introduction) 器的观测方程为一个增广的观测方程.这种方法可