Low Cost IMU Attitude Estimation Algorithm Based on Measurement Adaptive Reduced DCM

Attitude estimation is critical for navigation and control systems, and low-cost inertial measurement units (IMUs) have become a preferred choice for many applications due to their affordability, compact size, lightweight, and long lifespan. However, traditional attitude estimation algorithms, such as complementary filtering (CF) and quaternion-based extended Kalman filter (EKF), often perform poorly when the carrier is subject to nongravitational accelerations or external disturbances. This article presents a measurement adaptive reduced direction cosine matrix (MA-RDCM) attitude estimation algorithm. The algorithm constructs an EKF using a reduced direction cosine matrix as the state vector, integrating measurement data from MEMS gyroscopes and accelerometers for real-time attitude estimation. Furthermore, a measurement adaptive adjustment strategy is proposed, which dynamically corrects the measurement noise covariance matrix in real-time to address the degradation in attitude estimation accuracy caused by nongravitational accelerations during carrier motion. The adaptive adjustment strategy effectively reduces the impact of nongravitational accelerations and external disturbances, improving the accuracy and robustness of attitude estimation for low-cost IMU. The performance of the algorithm is evaluated through computer simulations and land vehicle experiments. The results show that the MA-RDCM algorithm significantly enhances attitude accuracy, particularly in dynamic scenarios, improving performance by approximately 33% compared to traditional methods, showcasing its significant advantages.