多分辨率传感器多尺度融合估计算法

Multi-Scale Fusion and Estimation Algorithm for Multi-Resolution Sensors

该文在基于信号统计特性的离散小波变换理论(DWT)和基于状态转移模型的动态多尺度系统理论(DMS)的基础上,提出了一种多分辨率传感器的多尺度状态融合估计新算法。该方法利用离散小波变换,首先对不同分辨率传感器Kalman滤波模型的状态方程和观测方程分别进行多尺度处理,构建统一的多尺度Kalman滤波模型,然后将不同分辨率传感器在同一尺度上获得的观测向量融合滤波,获得了优于已有多尺度状态融合估计方法的处理效果。并利用Monte Carlo仿真验证了该算法的有效性。

On the basis of theory of Discrete Wavelet Transform(DWT) of signal statistical characteristics and Dynamic Multi-scale System(DMS) of state transition model, a new algorithm for multi-scale fusion and estimation of multi-resolution sensors is derived. In order to construct a uniform resolution model of Kalman Filter,the equations of state and measurement are processed with DWT at different resolution levels, and then the measurements of the same resolution level are fused and filtered. Experimental results indicate that the proposed method is more effective than some existing algorithms of multi-scale fusion and estimation.