单频双差LM-BP神经网络GNSS周跳检测与修复研究

Research on Cycle Slip Detection and Repair of GNSS Based on Single Frequency Double Difference LM-BP Neural Network

针对传统周跳检测和修复方法精度低的问题,基于LM-BP神经网络理论,以全球导航卫星系统(GNSS)的载波相位双差序列为输入数据集,提出了一种周跳检测和修复方法。实验结果表明:对于小周跳探测问题,基于双差检测量的LM-BP神经网络探测修复法具有高敏感性,且相较于传统的多项式拟合法,新方法的周跳探测精度得到了提升。

In view of the low accuracy of traditional cycle slip detection and repair method,based on Levenberg Marquardt-Back Propagation(LM-BP)neural network theory,taking the carrier phase double difference sequence of Global Navigation Satellite System(GNSS)as an input dataset,a new cycle slip detection and repair method is proposed.The experimental results show that for the detection problem of small cycle slips,the LM-BP neural network detection and repair method based on double difference detection has high sensitivity.And compared with the traditional polynomial fitting method,the cycle slips detection accuracy of the new method is improved.