基于小波和RBF神经网络的多传感器时间对准算法研究

Algorithm of Alignment of Multi-Sensor Time Based on Wavelet and RBF Neural Network

提出利用小波降噪和RBF神经网络进行时间对准算法,该方法先对采集到的数据进行降噪处理,为下一步操作提供准确数据;然后利用RBF神经网络将不同传感器采集不同采样周期的数据对应到同一个时间点上,进行多传感器时间对准,为后面数据特征提取和数据融合提供高精度数据支持。通过仿真结果与其它方法的对比可得该方法运算速度快,具有数据修正能力和精确度高的特点,适用于多传感器的时间对准。

It proposed the use of wavelet reduces noise and RBF neural network for time alignment algorithm, first, this way can reduce the collected noise of data, as providing accurate data to next step; and then sampling period corresponds to the same point in time of different sensors collect data by RBF neural different networks, to the back data feature extraction and data fusion to provide high-precision data support by multi-sensor time alignment. The simulation results compared with other methods, the method computing is speeder, capacity of correcting data and higher accuracy, suitable for multi-sensor time alignment.