Adaptive Extended Kalman Filter Based on Genetic Algorithm for Tightly-Coupled Integrated Inertial and GPS Navigation

An adaptive extended Kalman filter is derived for integrating inertial measurements from gyros, accelerometers with GPS pseudorange and pseudorange rate measurements. The adaptive filter uses an estimator based on state residual to provide a positive definite estimate of the process noise covariance matrix. The genetic algorithm is utilized to optimally determine the estimator's parameter which is a slide window size. The filter formulation is based on standard inertial navigation equations. Simulation results are shown to compare the performance of nonadaptive and adaptive Extended Kalman Filter.