Print

Montane, Paul

Electrical Engineering

2009

MS

The objective of this research has been to achieve autonomous navigation for a hovercraft vehicle by designing and implementing an Inertial Navigation System (INS), thus making it capable of independently moving to a pre-specified set of GPS waypoints. The control of the craft involves fusing the data from a variety of sensors to make quick decisions on both obstacle avoidance and navigation. The primary sensor on the vehicle is a tri-axial orientation sensor, which serves as the Inertial Measurement Unit. The real-time data from this orientation sensor forms the inputs to a Simulink-based control model. A GPS sensor serves as an aid to the orientation sensor, by providing periodic correctional updates with regard to the position information. Collision detection is accomplished by using ultrasound rangefinders placed around the perimeter of the hovercraft's deck. A Kalman Filtering methodology combines the data from the sensor inputs. The overall strategy for the control system was to approximate the hovercraft mechanics with linear mathematical models. A small physical model that mimics the dynamics of the hovercraft tested the effectiveness of the Inertial Navigation control system that was developed.