论文部分内容阅读
不同的传感器数据采集系统采集的数据具有不同的分辨率,因而需要解决多分辨率数据的融合技术和方法。Kalman滤波对非平稳信号具有较强的估计能力,能对信号所有的频率成分同时进行处理;小波变换的多分辨分析正好提供了有效的多分辨率信息处理方法。因此本文基于小波变换的分时分频多分辨率特点,把信号进行小波变换,然后分别在各尺度上进行Kalman滤波估计,最后通过Mallat快速重构算法,得到融合后的结果。
Data collected by different sensor data acquisition systems have different resolutions, and thus need to solve the fusion technology and method of multi-resolution data. Kalman filtering has strong estimation ability for non-stationary signals and can process all frequency components of the signal at the same time. The multi-resolution analysis of wavelet transform provides an effective multi-resolution information processing method. Therefore, based on the time-division multi-resolution feature of wavelet transform, this paper carries out wavelet transform on the signal, and then performs Kalman filter estimation on each scale separately. Finally, Mallat fast reconstruction algorithm is used to obtain the fused result.