GTXR Kalman Filter Implementation
01. Overview
Implemented multivariate Kalman Filter in C for state estimation in GTXR flight computer, integrating a complementary filter for acceleration and angular velocity resulting in decreased estimation errors. Created Java matrix operations library to enhance understanding of Kalman Filter mathematics, with data handling in both Euler angle and quaternion formats.
02. Deep Dive
In aerospace and robotics, precise estimation of orientation and position is paramount. For the GTXR flight computer, I developed a highly optimized multivariate Kalman Filter in C. This implementation accurately fuses sensor data to provide a robust estimate of the vehicle's state, even in the presence of noise and sensor inaccuracies. To complement the core C implementation and deepen my own mathematical understanding, I authored a comprehensive matrix operations library in Java. The project rigorously handles data in multiple spatial representation formats, including both classic Euler angles (which are prone to gimbal lock) and quaternions (which provide stable, continuous rotational calculations). This dual-language approach allowed for rigorous testing and simulation in MATLAB and Java before final deployment in the embedded C environment.
Project Info
Achievement
Applications of Robotics
Timeline
Aug 2024 - Dec 2024