Abstract
The integration of the global navigation satellite system (GNSS) and the inertial navigation system (INS) is a well-established method for achieving accurate positioning, especially in applications involving unmanned aerial vehicles (UAVs). UAVs are increasingly used across various fields, yet they face challenges such as the need for real-time processing and the impact of low-quality measurements from cost-effective devices. To address these challenges, we propose a velocity-constrained, enhanced, real-time, low-cost, GNSS/INS integrated navigation algorithm and design an algorithmic platform based on the open-source software KF_GINS. The algorithm supports loosely coupled integration of GNSS position data and raw inertial measurement unit (IMU) data, utilizing a 4G data transmission unit (DTU) for real-time data transmission and performing loosely coupled computations on the received data. Subsequently, we successfully applied this algorithm to low-cost integrated navigation devices, such as UAVs. We tested the algorithm platform using one set of vehicle-mounted data and six UAV datasets. Experimental results indicate that the algorithm platform effectively performs computations under various conditions, improving single-point positioning (SPP) accuracy by up to 15.38% horizontally and 6.78% vertically. These findings demonstrate the algorithm platform's capability to significantly enhance the accuracy and stability of integrated navigation positioning for UAVs.