Robust longitudinal rate gyro bias estimation for reliable pitch attitude observation through utilization of a displaced accelerometer array.

摘要

In this thesis, a novel attitude estimation device is proposed utilizing cost-effective measurement sensors. The device fuses a rate gyroscope with an accelerometer array to estimate and eliminate the rate gyro bias online yielding accurate real time aircraft attitude tracking. Attitude determination algorithms are dependent on instantaneous and accurate measurements of translational and rotational body rates for precise estimation of vehicle orientation in three-dimensional space. Measurement error of instantaneous rate sensors, gyroscopes, is introduced via inherent biases and signal noise resulting in gyro drift. Integration of the rate signal for calculation of a net displacement amplifies these minute measurement errors leading to inaccurate and unreliable attitude observation. The proposed device is a departure from typical attitude observers and bias estimators due to its reliance on accelerometers measuring the local gravitational vector in lieu of additional magnetic field sensors or GPS. The end result of this work is a longitudinal attitude estimation device able to compute a rate gyro bias in real-time producing accurate pitch angle tracking while subjected to simulated aircraft flight conditions. The effectiveness of the newly constructed attitude estimation algorithm is demonstrated by comparison of attitude and rate gyro bias estimates produced from noise corrupted and biased sensors with the actual attitude of a nonlinear aircraft model and true rate gyro bias.