Fault detection and system re-configuration in flightcontrol systems based on optimization algorithms

In this thesis study, a robust, efficient, and optimized fault detection method has been developed for Automatic Flight Control Systems (AFCS). The main objective of the study is to detect possible faults that may occur in flight control systems with high accuracy while minimizing the false alarm rate. In this context, the Artificial Immune System (AIS), developed based on the Real-valued Negative Selection Algorithm (RNSA) and V-detector algorithms, was optimized using Genetic Algorithms (GA). Additionally, to improve the algorithm's anomaly detection performance and reduce the computational costs in real-time flight data monitoring, a hybrid approach was developed by integrating the Artificial Bee Colony (ABC) algorithm with the Negative Selection Algorithm (NSA). The AIS-GA and ABC-NSA software developed in this study were implemented in the MATLAB R2023b environment using real flight data obtained from a Boeing 737 aircraft belonging to a local airline company to detect potential faults within the flight control system. Faulty data were randomly generated to evaluate the software's performance. The results demonstrate that ABC-NSA approach outperforms the others, achieving the highest Figure of Merit (FoM) value with a 99.18% fault detection rate, a 1.11% false alarm rate, and an optimal detector count of 1000. When a fault occurs in flight control systems, this approach facilitates the rapid implementation of anomaly detection and identification, and contributes to enhancing flight safety and operational reliability by offering an innovative, low-cost, and computationally efficient solution for integration into real-time flight control systems.