This example shows how to design a fault detection, isolation, and recovery (FDIR) application for a pair of aircraft elevators with redundant actuators. The fault detection control logic used in this model is the same that is used in the Avionics subsystem of the Aerospace Blockset™ example entitled "NASA HL-20 with FlightGear Interface."
Figure 1: Top level of aircraft elevator control system model
A typical aircraft has two elevators attached on the horizontal tails (one on each side of the fuselage). There are a number of redundant parts in the system to enhance safety of the aircraft.
Figure 2: Schematic showing how the components of the elevator system are connected to one another
For example, as shown in Figure 2, there are:
Two independent hydraulic actuators per elevator (four total)
Three separate hydraulic circuits to drive the actuators
Two primary flight control units (PFCU)
Two control modules per actuator: full range control law and limited / reduced range control law
Figure 3: Actuator mode logic for fault detection
Each outer actuator has a dedicated hydraulic circuit, whereas the inner actuators share one hydraulic circuit. Each actuator can be in either one of five modes: Passive, Standby, Active, Off, and Isolated. By default, the outer actuators are in the Active mode, and the inner actuators are in the Standby mode. If a failure is detected in the outer actuators or in the hydraulic circuits that are connected to them, the fault detection system responds accordingly to maintain stability by turning the outer actuators off and activating the inner actuators.
If the aircraft is flying perfectly level, then the actuator position should maintain a constant value. If the position of an actuator increases or decreases by 10 cm from this zero point, then the fault detection system registers a failure in that actuator. The fault detection system also registers a failure if the change in actuator position is very rapid (i.e., the position changes at least 20 cm in 0.01 seconds).
Similarly, the fault detection system registers a fault in one of the hydraulic circuits if the pressure is out-of-bounds or if the pressure changes very rapidly. In this example, the fault detection system checks that the pressure in the hydraulic circuit is between 500 kPA and 2 MPa, and that the pressure changes no more than 100 kPa in 0.01 seconds.
Figure 4: GUI used to inject failures into the system
A separate GUI (shown in Figure 4) that was created in GUIDE is used to inject failures into the fault detection systems for both the hydraulic circuits and the actuators. By checking or unchecking boxes and pressing the Update button on the GUI, custom MATLAB® code runs as an intermediary between the GUI and the Simulink® model. For example, when checking the H1 box and pressing the Update button to inject a failure into hydraulic circuit 1, the following custom MATLAB code is evaluated:
% Define H1_fail Constant block location and get current value
blockname=[mname '/Signal conditioning '...
'and failures /Hydraulic Pressures/Measured ',char(10),...
'Hydraulic system 1 ',...
% Change value of H1_fail Constant block from 0 to 1 or from 1 to 0.
This effectively turns on a switch within the Signal conditioning subsystem so that the fault detection system registers a a fault in hydraulic circuit 1.
Figure 5: Switch that injects a failure into hydraulic circuit #1 if value of Constant block H1_fail does not equal 0
Stateflow® responds to failures in the hydraulic circuits and actuators using truth table functions, event broadcasting, and control logic. As an example, if the fault detection system registers a failure in hydraulic circuit 1 and no other failures occur, the L_switch truth table function evaluates Decision D1 as true and performs Action 2, which is to turn off the left outer actuator. This event is broadcast to the LO state, where the left outer actuator turns off. After this occurs, an event is broadcast from the LO state to the LI state so that the left inner actuator is activated. The right inner actuator is then activated since the left inner actuator is also active. When this happens, an event is sent from the RI state to the RO state to place the right outer actuator on standby. Thus, after the fault detection systems registers a failure in hydraulic circuit 1, the left outer actuator is turned off, the right outer actuator is placed on standby, and the inner actuators are activated.
When the fault detection system registers a failure occurs in one of the actuators, that actuator can no longer be activated. This is represented in the state chart by adding an Isolated state that contains no outgoing transitions. Therefore, once an actuator enters the Isolated state, it remains in the Isolated state.
A recovery mechanism has also been placed in the fault detection, isolation and recovery control logic in case a failed system comes back online. For example, if the fault detection system no longer detects failures in hydraulic circuit 1, the condition !u.low_press is true, and the LO state can transition from the Off state to the Standby state. That way, if the fault detection system registers a failure occurs in another system, such as the left inner actuator, then the left outer actuator can be activated.
Pieter J. Mosterman and Jason Ghidella, "Model Reuse for the Training of Fault Scenarios in Aerospace," in Proceedings of the AIAA® Modeling and Simulation Technologies Conference, CD-ROM, paper 2004-4931, August 16 - 19, Rhode Island Convention Center, Providence, RI, 2004.
Jason R. Ghidella and Pieter J. Mosterman, "Applying Model-Based Design to a Fault Detection, Isolation, and Recovery System," in Military Embedded Systems, Summer, 2006.