18 / 2023-03-26 11:56:44

Tight Integration of Low-cost GNSS PPP/INS for Seamless Land Vehicle Navigation

PPP; INS; tight integration; land vehicle navigation

This work presents a low-cost Precise Point Positioning (PPP)/Inertial Navigation System (INS) solution for land vehicle positioning in various environments including dense urban areas, underground parking, etc., which can be applied in land vehicle navigation, underground positioning, etc. In recent years, with the improved signal quality, low-cost Global Navigation Satellite System (GNSS) receivers are widely applied in various applications.  The presented PPP/INS system adopts u-blox M8U, a low-cost single-frequency GNSS and Micro-Electro-Mechanical Systems (MEMS) module, to provide a continuous solution for land vehicles. In an open-sky environment, with received State Space Representation (SSR) corrections, the system can provide a sub-meter PPP solution in most areas. In total blockage areas (i.e., underground parking, tunnel, mine, etc.), the INS with the MEMS inertial measurements can continuously provide the for land vehicles. Effective zero velocity detection and application of Zero Velocity Potential Update (ZUPT) together with velocity constraints of land vehicles effectively reduce the quick drift of INS caused by the accumulation of low-cost MEMS errors.  The most challenging scenario for the PPP/INS system is areas like dense urban where the positioning error could be up to tens of meters due to the abnormal GNSS measurements caused by multipath and none-line-of-sight (NLOS) receptions. The robust positioning in GNSS challenging environment is of great importance since severe GNSS signal blockage is common in an urban canyon, roads with trees, etc. Besides, before entering a tunnel or mine, GNSS could be partially blocked, leading to large positioning errors. As a result, the Dead Reckoning (DR) results of the INS system would drift if large errors exist in the initial state. The presented PPP/INS system applies a robust outlier detection method with aid of inertial measurements, in which a prime satellite is selected based on noise level, signal strength, etc. from each constellation to reject the outliers with the combination information of INS mechanization, satellite orbit, clock, etc. The general performance in terms of availability and accuracy can be improved by detecting and rejecting the GNSS outliers. In summary, the presented PPP/INS system could provide seamless positioning for land vehicles in most practical applications with low-cost sensors, which makes it accessible to a wider range of vehicles.