Abstract:

In the following thesis, a Kalman filter is applied to estimate the state of an aircraft. The states are attitude, position, gyro bias and accelerometer bias. The Inertial Navigation System is used as the primary navigation system, and GPS and line of sight provide measurements to be used in the overall system. The performance of the filter includes how well the filter estimates the state errors as well as speed of convergence. Several cases are studied, varying initial conditions and number of measurement sources. The only state variable treasured is position in Earth-Centered-Earth-Fixed coordinates. The other states are coupled with position and hence allows us to estimate them "reasonably". It is observed that the filter is sensitive to initial conditions. The filter works best with the initial conditions closer to true values. The filter does not perform well when only GPS measurements are used along with "large" initial condition errors. However, the filter is able to handle "large" initial condition errors with some values outside the 3? bound with the addition of line of sight measurements.