Abstract
The real-time, high-precision determination of a moving object's geological position (latitude, longitude, altitude) is a critical challenge in fields such as autonomous navigation, precision agriculture, and geophysical surveying. While Global Navigation Satellite Systems (GNSS) provide the primary data source, raw signals are susceptible to errors from atmospheric delays, multipath effects, and receiver noise, leading to positional inaccuracies. This article presents a monitoring system that leverages an advanced mathematical framework based on an Adaptive Unscented Kalman Filter (AUKF) to fuse multi-sensor data and achieve centimeter-level precision. Building directly on our previous work in robust filtering and error-specific modeling, we develop a non-linear state-space model that incorporates advanced tropospheric and ionospheric delay corrections. The AUKF core dynamically estimates and compensates for state and measurement uncertainties in real-time. Simulation results demonstrate a significant improvement in positional accuracy and robustness compared to standard Extended Kalman Filter (EKF) approaches, particularly during periods of high dynamics and signal degradation, validating the concepts proposed in our earlier theoretical research.