Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.
Issue link: https://insideunmanned.epubxp.com/i/720234
55 unmanned systems inside August/September 2016 ugust/Septe August/September 2016 August/September 2016 be August/September 2016 0 August/September 2016 6 August/September 2016 ENGINEERING. PRACTICE. POLICY. The designed residual filter is based on this true servo model. It uses measured actuator data as inputs to generate a residual, which is sensitive only to actuator faults. The decision logic reads this residual signal and compares it with preselected thresholds to determine the occurrence of faults. There is a high design f lexibility for the residual filter and decision logic. The efficacy of the decision logic is usu- ally improved by analyzing the probability of false alarms and missed detections and adapt the parameter setting of the fault diagnosis system based on these probabilities. Once a fault has been recognized in the servo, the available fault information can be used in the fault tolerant control system to switch to a new controller and prevent loss of control of the UAS. Control System Reconfi guration: As mentioned in the preceding section, fault diagnosis in- volves detection, identification and estimation of the fault. This is followed with some action to mitigate the effects of the fault. In the event of control surface faults, the control input com- mands are reallocated to the remaining func- tional control surfaces and the flight control law is suitably reconfigured. However, the worst- case scenario is when the aircraft is down to its last functional control surface and there are no remaining surfaces that can be reallocated. The use of all aerodynamic control surfaces is necessary for maximizing f light performance. However, one control surface is sufficient for the aircraft to safely fly home. In the winter of 2015, we conducted several f lights demonstrating that an aircraft can be safely f lown with only one functional aerody- namic control surface. This was to show that a UAS with all but one failed control surface servo can be landed safely. The technology demonstration platform was a modified Ultra Stick 120. The modification was slight and is shown in Figure 4 and its objective was to in- corporate a modicum of hardware redundan- cy. The modification consisted of: (1) Splitting the single rudder into two independently ac- tuated control surfaces and (2) Breaking the mechanical linkage between the two elevator surfaces and allowing them to be actuated in- dependently using two separate servos. The final reconfigured Ultra Stick 120 had a total of eight aerodynamic control surfaces: a pair each of elevators, ailerons, rudders, and f laps. These control surfaces are labeled in Figure 4. For the demonstration, only the left elevator (labeled E1) was actively commanded. The re- maining seven control surfaces were frozen at their respective trim positions. In addition, the Figure 4. The modifi ed Ultra Stick 120 aircraft. Figure 3. Model based fault diagnosis setup for a servo actuator in a UAS. Figure 5. The Goldy fl ight computer.