Application of artificial intelligence methods for the designand development of aircraft flight control laws

The development of flight control systems has always been the main topic of many studies. In commercial aviation, it is essential to address the aircraft stability, robustness, and precise tracking performance for ensuring passengers safety and comfort during the entire flight, especially during the cruise phase and in different flight environments. While conventional controllers revealed accurate performance, integrating Artificial Intelligence (AI)-based methodologies can offer new features and capabilities to shape the next generation of flight control systems. These controllers can be developed without explicit knowledge of the aircraft model, effectively handle uncertainties, and control the aircraft with a fixed parameter configuration for all flight conditions with enhanced adaptive characteristics. The first article discusses a control methodology constructed by a Type One Adaptive Fuzzy Logic System (T1AFLS) and a Sliding Mode Control system (SMC), to control the aircraft pitch rate and True AirSpeed (TAS) during cruise. The T1FLS approximates the unknown dynamics, which are updated using adaptation laws designed by the Lyapunov theorem. Subsequently, the approximated functions are integrated into the SMC system to guarantee the aircraft tracking performance and ensure its stability and robustness. In the second article, the T1FLS was converted to an enhanced Type Two FLS using a new Type Reduction algorithm. This two-dimensional approximator can handle a larger number of uncertainties. The simulation results for this combination of the T2FLS, Adaptive Control, and SMC systems were compared with those of the T1AFSMC system. The comparison revealed that the T2AFSMC performed slightly better than the T1AFSMC. Both T2AFSMC systems, employed for both pitch rate and TAS control systems, could meet the requirements of aircraft stability, robustness, and tracking performance during the cruise. The third study focused on developing a control system for the aircraft lateral motion. The T2AFLS was employed to approximate the unknown aircraft dynamics during flight. This approximator was employed within a super-twisting sliding mode control system, enhanced with adaptation laws, and with the Particle Swarm Optimization (PSO) algorithm to find parameter values of the sliding mode controller. To address the coupled roll and yaw dynamics modes, an Integral controller acting on the aircraft sideslip angle, was used to stabilize the yaw rate indirectly. These methodologies were evaluated to ensure the aircraft appropriate performance, and the Adaptive super-twisting sliding mode performed better than the PSO-based sliding mode control. In the fourth paper, an autopilot control system was developed using a combination of Fuzzy Recurrent Neural Network (FRNN) and SMC systems. In this autopilot system, two separate FRNNs were developed for both Vertical Speed and Altitude Hold modes, which dynamically approximate the aircraft dynamics. Moreover, two sliding mode controllers were applied to allow the aircraft to track the reference signals in each mode. This autopilot system employed a new fuzzy transition algorithm to switch between these modes. The proposed methodologies were validated by a nonlinear simulation platform, developed using flight data from a Level D research aircraft flight simulator for the Cessna Citation X aircraft at the LARCASE in different flight conditions.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.