针对传统小波在惯性传感器去噪中存在计算复杂和分解层数固定的问题,提出一种基于模糊逻辑的提升小波阈值去噪方法.该方法首先利用模糊逻辑判断载体当前的运动状态,并根据先验知识确定该状态下的信号带宽;然后依据小波分解层数与截止频率间的对应关系,选择惯性传感器各个轴向的分解层数;最后在提升小波框架下对信号做离散小波变换,并对各层系数进行阈值去噪.实验结果表明,该方法可以对运动状态进行正确判别,与固定层数的传统小波相比,具有更好的降噪效果和更快的处理速度.%As the classical wavelet had the shortcoming of complex computation and fixed number of decomposition layers in de-noising of inertial sensors, a threshold de-noising method based on fuzzy logic and lifting wavelet transformation was proposed. Firstly, the motion state was determined with a fuzzy expert system, and the signal bandwidth of the true motion state was chosen accord to the spectral estimation based on a priori knowledge. Then, suitable wavelet decomposition layers for each axis of an inertial sensor were automatically chosen in accordance with the relationship between wavelet decomposition layers and cut-off frequency. Finally, a discrete wavelet transformation was conducted under a lifting wavelet frame, and the signal was de-noised by compressing the coefficients of wavelet transformation of the signal at a threshold value. Test results proved that the proposed de-noising method can distinguish the motion state correctly, and has a better results and higher computational efficiency than the classical wavelet de-noising with the fixed number of decomposition layers.
展开▼
