Simulation and Control of Quadplanes: a Particle Swarm and Model Predictive Control Approach

This thesis introduces a comprehensive flight control system optimized through evolutionary algorithms. This involves developing a quadplane simulator using Simulink in collaboration with Sky Eye System company to simulate and control quadplanes, vehicles that combine VTOL capabilities with fixed wing endurance. The simulator is central for testing and enhancing the control strategy, particularly in the transition phase, from hovering mode to fixed wing mode and vice versa. The work focuses on the challenges of the transition phase, aiming at achieving a smooth transition, without excessive maneuver, particularly during landing. A significant portion of the work is dedicated to fine-tuning PID controllers using Particle Swarm Optimization (PSO), which yields substantial improvements in stability, efficiency, and precision tracking. The robustness of the control system is tested against model uncertainties and varying wind conditions. The latter part of the study explores Model Predictive Control (MPC) as a potential future implementation. Initially applied to the quadcopter mode, this exploration brings to light the challenges associated with MPC, highlighting the importance of developing an efficient predictive model for enhanced performance.