A Finite-Time Complementary Filter for Attitude Estimation Using Inertial Sensors

This paper presents a novel finite-time complementary filter specifically designed for attitude and gyroscope bias estimation. The key contributions of this filter include ensuring rapid convergence to the actual attitude within a finite time, significantly improving performance over traditional filters that only guarantee asymptotic stability. The filter achieves high accuracy by effectively integrating data from multiple sensors, such as gyroscopes, accelerometers, and magnetometers, through advanced sensor fusion algorithms and noise suppression techniques. Additionally, simulation tests confirm the filter's robust resistance to noise, highlighting its potential to enhance the performance of attitude estimation systems across various applications by providing faster, more accurate, and reliable estimates.