Bio-Inspired Optimization of Fractional-Order Control in Drone Systems

Abstract

Robotics has become a deeply interdisciplinary field, addressing a broad spectrum of challenges to support human activities across diverse environments on land, in the air, underwater, and even in space. This study focuses on the quadcopter, offering a detailed analysis of its modelling, control strategies, and performance optimisation, particularly when navigating sharp trajectories. Stability plays a crucial role in the effectiveness of quadcopters and other robotic systems. In this context, the research explores how a fractional-order PID (FOPID) controller can enhance stability, especially in systems characterized by high sensitivity. To achieve optimal control, five fractional parameters of the FOPID controller are tuned using three distinct optimisation techniques: Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Bonobo Optimization (BO). The comparative evaluation reveals that the FOPID controller, when optimized particularly with the BO method, demonstrates superior control performance. It notably eliminate overshoot and improves both settling and rise times by more than 30% under consistent test conditions, outperforming conventional controllers cited in the literature.