Precise Orientation Estimation Based on Nonlinear Modeling and Quaternion Transformations for Low Cost Navigation Systems

An adaptive unscented Kalman filter (AUKF) is designed to estimate the roll and pitch angles of rigid body using low-cost inertial sensors. The main challenge is concerned about the tilt orientation estimation in high dynamic environments, where the linear acceleration affects the orientation accuracy. The proposed filter is based on the quaternion technique and an additive function which is used to compensate the influence of accelerometer corrections during motions. The algorithm is implemented on a STM32F407 ARM Cortex-M4 series microcontrollers and fused three-axis accelerometer and, three single-axis gyroscopes triads based on micro-electromechanical system (MEMS) technology. The experimental and field tests results analysis showed an outstanding real-time navigation performance when compared with the traditional KF and other commercial expensive systems.