A new method of terrain self-adaptive matching algorithm for autonomous underwater vehicle

As different underwater terrain features will affect the accuracy of unscented Kalman filter terrain matching algorithm, a new terrain self-adaptive matching method is proposed which adjusts sigma point distribution by terrain features. The connection between sigma point distribution distance and three basic terrain features were analysed. Positioning error range were calculated using navigation positioning errors and underwater digital map. Terrain elevation standard deviation was used to characterize the information quantity of decision region. Scaling parameter was adjusted using linear mapping method and sigma point distribution distance was determined by terrain features. The simulation results proved better terrain adaptability and positioning precision of improved self-adaptive matching algorithm.