A LiDAR–INS-aided geometry-based cycle slip resolution for intelligent vehicle in urban environment with long-term satellite signal loss

Intelligent vehicles usually equip with GNSS receivers and MEMS-IMU for localization and light detection and ranging (LiDAR) sensors for perception. Cycle slip detection and repair is significant for the GNSS receivers to achieve high-precision positioning results. We propose an improved geometry-based cycle slip detection and repair method considering the positioning error at the immediate prior epoch and the influence of the satellite geometry change. Experimental results show that our improved method can improve fixed rates of ambiguity resolution compared to the traditional geometry-based method, especially for long-term satellite signal loss. Based on the improved method, we further propose a LiDAR–INS aiding LAMBDA method for cycle slip detection and repair. The INS model and LiDAR scan-to-map matching results are fused by extended Kalman filter (EKF) to calculate between-epoch relative position, which provides a constraint for LAMBDA. Experimental results prove that our LiDAR–INS positioning method can get centimeter-level accuracy for small observation gaps (e.g., 10 s) and decimeter-level accuracy for large observation gaps (e.g., 1 min), which can help achieve high fixed rates and success rates (e.g., 0.85) for large observation gaps (e.g., 200 s) in the urban occlusion environment where prior epoch owns meter-level positioning accuracy.