On-line scheduling and fault detection in NCS with communication constraints in Drone application

This paper considers the fault detection problem of mini-helicopter (Drone) where sensing and actuation signals are exchanged among various parts of the system via communication networks. For reducing the network load and thus minimizing the uncertainty caused by transmission delay or packet dropout, instead of transmitting all measurements of velocity sensors at each sampling time, only one velocity sensor can have access to the network to transmit its measurement. In order to increase accuracy and rapidity of fault detection in motor of Drone, an approach based on Multi Sequential Probability Ratio Test for residual classification and fault detection of each motor is proposed. Then in function of accepted hypothesis of each sequential test a new communication sequence is chosen from ensemble of admissible communication sequences which preserves more medium access to a sub-system (motor) with more deviation from its normal condition.