An Intelligent Rule-Based System For Fault Detection And Diagnosis On A Model-Based Actuator Device

Unmanned aerial vehicles due to their large operational potential may be required to travel over long distances and through various weather conditions, which might lead to potential degradation or even failure of their electrical or/and mechanical actuator parts. Control in trajectory derivations and path following processes is highly dependable on these actuators and sensors. Depending on their efficiency, the outcome will be a near optimum solution to every problem. Consequently, the minor failure can degrade the performance of the process and might drive it to an uncontrollable system. Therefore, an efficient mechanism should be capable of making these faults realizable and act accordingly so that a consistent performance actuator performance qualitative or quantitative index is continuously maintained. In this paper electro-mechanical actuator potential failures are firstly detected and then diagnosed for the application of unmanned aerial vehicles. It includes several scenarios of actuator faults and results which demonstrate the fault conditions and the effectiveness of the detection and diagnosis Kalman based algorithms. It involves the diagnosis strategy to minimizing errors produced due to malfunction in components or inaccuracies in the model. The residuals used are generated using empirical actuator models which are chosen under specific operating regimes.