Experimental evaluation of data fusion algorithm for residual generation in detecting UAV servo actuator fault

No Thumbnail Available
Sahwee Z.
Rahman N.A.
Sahari K.S.M.
Journal Title
Journal ISSN
Volume Title
Multi-Science Publishing Co. Ltd
Research Projects
Organizational Units
Journal Issue
Unmanned Aerial Vehicles used by military are generally designed with very high reliability and have multiple redundancy of software and hardware equipment because they are intended to operate in hostile environment. But, relatively low cost UAVs used commercially are not equipped with such systems. Usually, micro UAVs weight less than 2kg are equipped with on-board miniature sensor and operate without any hardware redundancy and thus could reduce their reliability. Some of these commercial UAVs that operate in populated areas will cause damage and fatality if faulty system occurred. Hence there is a need for on-board fault detection and isolation system without degradating the UAV flyability and its cost. Analytical redundancy or model reference method of fault detection algorithms could be implemented as most UAVs are microcontroller controlled. Together, with the availability of miniature sensors could provide an ideal platform for implementing fault detection. In this paper, the development of fault detection through residual generation algorithm is implemented with data fusion from miniature sensors. Some of these sensors are already installed within the autopilot system which reduce the amount of additional sensors needed. Identification of fault in the elevator is simulated experimentally and fault detection rate is monitored. From the implemented algorithm, the data fusion from additional sensors shows improvement in fault detection rate.
Algorithms; Data fusion; Hardware; Military vehicles; Redundancy; Software reliability; Unmanned aerial vehicles (UAV); Analytical redundancy; Data fusion algorithm; Experimental evaluation; Fault detection algorithm; Hostile environments; Model reference methods; On-board fault detection; Software and hardwares; Fault detection