Sensor fusion method for altitude estimation in mini-UAV applications

The emergence of unmanned aerial vehicles for large scale monitoring of ground infrastructure has challenged new designs for electronic sensor systems capable of timely, accurate and compntationally effective measurements. A conventional design issue is balancing the complexity of the sensing subsystem with advanced data processing algorithms in order to obtain quality information regarding the airplane's posture and orientation. The paper presents the generic architecture of the navigation sensing system for a fixed-wing mini-UAV, at both the physical and algorithm levels. The focus is set on a sub-problem, namely altitude estimation by sensor fusion using barometric pressure and GPS data. Results are discussed based on actual flight log data and the approach is extended towards speed estimation: air, ground and wind influence. Conclusions are drawn regarding efficient implementation of the method in relation to the onboard embedded processing capabilities.