加性噪声相关时带乘性噪声系统的最优估计-信号与信息处理专业论文.docxVIP

摘 摘 要 L研究带乘性噪声系统的信号估计理论是随机信号处理的一 个重要发展方向，在石油地震勘探、通讯工程、语言处理等应 用领域都具有实际意义。；本文围绕带乘性噪声系统的信号最优 估计问题展开了进一步的研究。 首先，本文回顾了带乘性噪声系统最优估计理论的发展， 在前人的工作基础上，就一种更一般的噪声条件提出了带乘性 噪声系统最优滤波算法，即就动态噪声为有色噪声及动态噪声 和量测噪声在有限时间段上相关的情形，给出了状态最优滤波 的直接算法及扩维算法。当对噪声的假设退化为非平稳独立白 噪声条件时，本文的算法与Rajasekaran滤波算法是等价的。 总结以往带乘性噪声系统的平滑估计理论和反褶积理论的 发展，就动态噪声为有色噪声且动态噪声和量测噪声在有限时 间段上相关的情形，在其状态最优滤波扩维算法的基础上，给 出了固定臂长平滑算法及固定域平滑算法。还进一步给出了递 推和非递推形式的固定域反褶积算法。 ／除在理论上给出上述各算法在线性最小方差意义下为最优 的证明之外，本文还给出了对应的仿真实例，验证了算法的有 效性。)、 关键词：乘性噪声?随机系统，相关噪声，线性最小方差? 最优估计，状态滤波与平滑，反褶积o＼ ABSTRACTResearching ABSTRACT Researching signal estimation theory for stochastic systems with multiplicatiVe noise is an important developing orientation in stochastic signal processing．It can be used in such fields as oil explo ration，communication engineering，target tracking and speech signals processing．In this dissertation，the important theoretical and practical problems of this theory，including optimal filtering，smoothing estimation and optimal deconv01ution estimation for system with multiplicative noise under the tendition that white measurement noise and C010red state noise are correlated on a finite time interval in the sense of linear minimum-variance are extensively diSCUSSOd． Fi rst，the development of Signal estimation theory for stochastic systemS with multiplicative noise is recalled in thiS dissertation． An optimal filtering algorithm for system with multiplicative noise under the condition that white measurement noise and colored state noi SO are correlated on a finite time interval in the sense of linear minimum。 varianee is developed． When state and measurement noises are ancorrelated white noises respectively，the algorithm is equal to Rajasekaran filtering algorithm． Based on the optimal filtering algorithm，two kinds of smoothing algorithm for the above system are developed．They are fixed-interval smoothing algorithm and fixed-1ag smoothing algorithm．The corresponding fixed·interval deconvolution algorithm i S advanced．It is presented