卫星信号失效时基于道路编码的路端定位方法

Road-end Assisted Vehicle Positioning Method Based on Road Code in GNSS-denied Environment

摘要:

在城市峡谷、隧道等GNSS信号易受干扰或完全失效的环境下，因缺少绝对定位信息，车辆定位精度会明显下降。针对这一问题，提出了基于道路编码的路端辅助定位方法。首先，设计了包含定位和校验信息的道路编码和对应解码方法，并利用编码规则实现了道路编码图案的自动生成；之后，设计了能够同时回归角点像素坐标和类型的道路编码角点检测网络，使得车辆能够准确、鲁棒地获得路端定位信息；最后，提出了一种GNSS失效环境下使用道路编码辅助车辆定位的参考范式，利用因子图优化算法将路端定位信息与车端定位信息融合。道路试验结果表明，该路端辅助定位方法能够显著提高车辆的定位精度和可靠度。

Abstract:

In environments， where GNSS is susceptible to interference or completely ineffective （such as urban canyons and tunnels）， the vehicle positioning precision will be significantly reduced due to the lack of absolute localization information. To address this problem， a road-end assisted positioning method is proposed based on road code. Firstly， the road code containing localization information and corresponding decoding method are designed， and the automatic generation of road code patterns is realized according to coding and verification rules. Secondly， a road code corner detection network that can simultaneously regress the corner pixel coordinates and type （whether it is the upper left corner point） is designed， so that vehicle can obtain road-end localization information accurately and robustly. Finally， a reference paradigm for using road code to assist vehicle positioning in GNSS-denied environment is proposed. The factor graph optimization （FGO） algorithm is used to integrate road-end localization information with that of vehicle-end. The real vehicle road tests show that the proposed method can significantly improve the vehicle positioning precision and reliability.