Attitude estimation of connected drones based on extended Kalman filter under real outdoor environments

In this study, we design connected multiple drones through electromagnets, and it requires an attitude estimation method that does not use magnetic sensors. In order to solve this problem, we propose an extended Kalman filter-based attitude estimation method based on the multiple global positioning system (GPS) to substitute the magnetic sensor. Additionally, we demonstrate that combining velocity vectors allows for highly accurate attitude estimation. We compare the estimation accuracy and stability of attitude estimation in simulation and real environments using two types of experiments. It is confirmed that the accuracy of yaw angle estimation of the proposed method is improved 36.1% in simulation and 29.7% in the real environments compared to conventional extended Kalman filter.

Automated summary

In this study, a connected multiple drone system using electromagnets is designed, and an attitude estimation method based on an extended Kalman filter and multiple GPS is proposed to replace magnetic sensors. Experimental results demonstrate that the proposed method achieves higher accuracy and stability in attitude estimation compared to conventional methods in both simulation and real environments.