Real Time Integration GPS with Gyro-Compassing Using Two Cascaded EKF with Free RTOS

Positioning autonomous vehicles accurately and reliably remains a crucial challenge, particularly in challenging environments where global positioning system (GPS) signals are weak or nonexistent. Integrated navigation systems (INS) with micro electrical mechanical system (MEMS) sensors offer a promising solution, but individual systems are afflicted with limitations in accuracy and susceptibility to environmental considerations. As a result, it is essential to construct an integrated system that combines INS with GPS. This system will provide a reliable solution to optimal positioning. This paper proposes a complete INS/GPS utilizing two cascaded Kalman filters to achieve the optimal navigation solution. The first filter, an adaptive extended Kalman filter (AEKF), fuses accelerometer and gyroscope data to obtain precise attitude angles, and heading angle computed from the magnetometer. The second filter is the extended Kalman filter (EKF) assists the INS velocity and position estimates with GPS data, further enhancing accuracy. This proposed integrated system is implemented on STM32F469 Cortex-M4 embedded system by using a free real time operating system (free RTOS) demonstrating its real-time capabilities on a land vehicle. The results have shown significant improvements in the execution time in comparison to standard EKF and better accuracy and robustness for positioning.