Fault Tolerant Control of a Missile Autopilot System
Özet
In aerospace defense industry, missile systems have been an important asset for more than half a century. As technological advancements take place, this field also continues to improve and develop today. A missile is simply a type of controlled aircraft with the purpose of carrying a payload (warhead) to a designated location. In order to complete this mission, a controller that can handle disturbances and uncertainties is necessary. The section responsible for the flight control of missiles is called the missile autopilot. The autopilot contains pre-designed controllers and other logical coding that concern any type of control of the missile. In this thesis, the main controllers of the autopilot are designed benefiting from model predictive control theory and then missile autopilot system is equipped with fault tolerance capability against actuator failure. Fault-tolerant control is a control approach that aims to recover controller performance in case of a fault occurring in the system. These faults can happen in the plant, sensors, or actuators. Fault-tolerant control takes place when the basic controller would fail in such a scenario. In missile systems, actuators are vital for the control of the missile. They rotate control surfaces of the missile so that the missile can carry out required maneuvers during flight. If one of the actuators beocme dysfunctional this can lead failure of the whole missile system. In order to prevent losing an entire expensive system due to actuator failure, an active fault tolerant control method is required. In this study, firstly the equations of motion for the system are obtained and then a linearization process is followed. Adaptive Model Predictive Control is applied to the linearized system model. The accuracy of linearization is shown by comparing time responses of linear and nonlinear system dynamics. Then a nonlinear analysis model is used to analyze the robustness and stability of the controller. Then the effect of a fault scenario is analyzed where one of the four actuators has failed and stuck at its latest position (also called total loss of effectiveness). Then a fault diagnosis logic is coded in order to locate and identify the faulty actuator. After the diagnosis, an appropriate fault-tolerant control process is applied to the remaining three actuators. Finally, nonlinear simulation results are obtained in order to compare the fault scenarios with and without fault-tolerant control and demonstrate the effectiveness of the method used.
