Fault detection, isolation and reconfiguration in navigation system of Boeing-747 aircraft

In this thesis; Boeing 747 aircraft mathematical modelling, INS, GPS, INS error modelling, Kalman Filtering, INS/GPS integration, fault detection, isolation and reconfiguration are widely explained. In the study, Kalman filter is used to estimation of Boeing 747 aircraft states. The INS error modelling and integration of INS/GPS are examined and Kalman Filter is used to estimate states for INS error and integrated INS/GPS models. Fault detection, isolation and reconfiguration on Boeing 747 aircraft model and integrated INS/GPS is done by Kalman filter. All these simulations are generated on MATLAB simulation environment. There are two common methods for navigation, one of them is Dead Reckoning and other one is Position Fixing. INS is a type of Dead Reckoning navigation and most preferable, while GPS is more popular in Position Fixing method. Global Positioning System (GPS) and Inertial Navigation System (INS) provide navigation solutions to the aircrafts, ships, spacecraft, missile and other vehicles. INS is comprised by navigation computers and inertial sensors such as accelerometers and gyroscopes. INS procures information about aircraft's positions, velocities and attitudes. Nowadays, different type of aircraft systems such as Flight Management System, Electronic Flight Instrument System, Flight Guidance and Control System use INS output data. The inertial navigation system have different type of errors such as initiliazation errors and inertial sensors errors. The main problem of the system is that these errors expand with time, so that INS is time dependent. INS gives high accuracy for short time because INS errors increase with time and system accuracy decreases. GPS is well-known system for navigation and it is used in different areas such as mobile phones, aircrafts, ships, automobiles and so forth. GPS is consisted of three different segments which are control, user and space. The user segment is GPS receiver and it calculates own position by using GPS signals which are send from GPS satellites on space. The control segment monitor and make correction to GPS data. Mostly used GPS has some type of errors due to different causes. For instance signal quality of GPS may decrease due to signal interference and navigation information may not be satisfied. As a consequence, stand-alone INS or GPS does not sufficient solution for navigation applications which need high precision such as aircraft navigation. These systems have pros and cons, so that these systems are integrated and their advantages are combined for better navigation solution. The integration of INS and GPS is carried out by Kalman Filter. Kalman filter has recursive data processing algorithm and the main features of the filter are integrates more than one systems, diagnose system noises and computes unknown variables and minimizes error of measurements. In this study there are different type of problems are solved by using Kalman Filter. These solutions are that the filter is used to estimate state variables of Boeing 747 aircraft, Kalman filter minimizes the measurement error such inertial errors and find best value of measured parameters, the filter make prediction for INS errors and it is used for integration of INS/GPS, and Kalman filter is used to fault detection, isolation and reconfiguration. Faults may occur any time in the systems, this affects system reliability and sustainability. Therefore, systems should be designed to minimizing fault effects. In this study, it is examined and observed that the filter is reliable and providing correct data, when faults happen. The aim of fault detection and isolation is that when fault is occured, fault should be successfully detected and isolated. In MATLAB simulation environment, faults are given to the systems which are integrated INS/GPS and Boeing 747 aircraft model. Kalman filter is used to fault detection, isolation and reconfiguration. Simulation results show that Kalman filter is very effective and reliable in faulty condition. When fault has occured in a system, fault is detected and isolated by using Kalman filter and the filter maintains correct estimation for states.