船舶微惯导网络中的数据采集及其小波降噪方法

摘要

微惯性器件普遍存在噪声大、稳定性差等问题，仅靠单一MIMU通常难以满足现代导航系统提出的精度要求。基于CAN总线，面向船舶实际应用，设计了一种采用MIMU多点布局的网络化微惯导系统。它主要由MIMU、数据采集、数据管理、时间基准、CAN智能转换器等模块以及工控机组成。各数据采集模块与工控机均以节点形式挂接于CAN网络，时间基准模块用于产生统一的时间刻度值及各通道MIMU报文同步采集信号，数据采集与数据管理模块之间采用中断握手方式来协调各通道报文上传的次序。最后，采用小波分层阈值法对MIMU动态实测数据进行了降噪处理。实验表明：该船舶微惯导网络系统能够基于CAN总线实现多通道MIMU实时数据的同步采集和网络化传输，误码率低；且采用小波分层阈值法对MIMU数据进行降噪处理是切实有效的。%All micro inertial components have common defects:large noise and poor stability, as the result, a single MIMU can not satisfy the precision demanded by modern navigation system. Based on CAN bus, a kind of networked micro inertial navigation system with the layout of multi-MIMU is designed for the application in marine. The system consists of MIMUs, data acquisition, data management, time reference, CAN intelligent converter and remote IPC. CAN bus is articulated with the nodes of all the data acquisition modules and IPC, the time reference module generates a uniform time scale value and synchronous acquisition signal of all MIMUs, the interrupt handshaking mode between the data acquisition and the data management modules is adopted to coordinate message uploading orders of all MIMUs. Finally, actual dynamic data of MIMU are denoised with the method of wavelet multi-level threshold. Experimental results show that based on CAN network, the system can ensure multi-channel real-time data of MIMUs to be collected synchronously and transmitted reliably with lower bit error rate. It is shown that wavelet multi-level threshold is effective in MIMU data denoising.