Adaptive Markov Model Analysis for Improving the Design of Unmanned Aerial Vehicles Autopilot

The need for Unmanned Aerial Vehicles (UAVs) is increasing as they are being used across the world for various civil, defense and aerospace applications such as surveillance, remote sensing, rescue, geographic studies, and security applications. The functionalities provided by the system is based on the system health. Monitoring the health of the system such as healthy, degraded (partially healthy or partially unhealthy) and unhealthy accurately without any impact on safety and security is of utmost importance. Hence in order to monitor the health of the system to provides the functionality for a longer period of time system fault detection and isolation techniques should be incorporated. This paper discusses Fault Detection and Isolation (FDI approach used in Unmanned Aerial Vehicle (UAV) autopilot to make its functionality more robust and available for a longer period of time. We proposes an integrated Adaptive Markov Model Analysis (AMMA) to detect and isolate faults in critical components of the system. The effectiveness of the novel approach is demonstrated by simulating the modified system design with FDI incorporation for the UAV autopilot. The proposed FDI approach helps in identifying the gyro sensor failure and provides a degraded mode to the system functionality which did not exist earlier in the design. The simulation demonstrates the system modes such as healthy, degraded (partially healthy or partially unhealthy) and unhealthy to understand the functionality better as the current design which works in only two modes i.e. healthy and unhealthy.