Advancing aerial robotics: a multi-strategy optimization framework for intelligent Hexarotor control
Hamza Tahiri, Ismail Mchichou, Mohamed Amine Tahiri, Mhamed Sayyouri, Eman Abdullah Aldakheel, Doaa Sami Khafaga · Swarm and Evolutionary Computation · 2026
This paper presents a Multi-strategy Quadratic Interpolation Optimization (MQIO) algorithm for precise tuning of fractional-order PID controllers in hexarotor aerial robots. The proposed method extends the original QIO framework by integrating a Generalized Quadratic Interpolation scheme, adaptive parameter scheduling, a Differential Evolution-inspired mutation operator, and Lévy-flight-based global exploration. This hybrid architecture effectively balances exploration and exploitation, mitigating premature convergence in complex, high-dimensional optimization landscapes. The algorithm is first rigorously evaluated on the CEC-2022 benchmark suite, achieving the best average rank and demonstrating statistically significant superiority over state-of-the-art metaheuristics, as confirmed by non-parametric Friedman and Wilcoxon tests. MQIO is then deployed to solve the core control challenge: automated tuning of a cascaded FO-PID controller for a nonlinear hexarotor model, minimizing an ITAE criterion under realistic parameter bounds. Simulation results on demanding 3D trajectories, a helical path and a back-and-forth motion under persistent external disturbances show that the MQIO-optimized controller delivers exceptional tracking accuracy and robustness. It attains the lowest recorded position RMSE (0.2134 m) and attitude RMSE (0.3151 rad) among all compared algorithms, including QIO, DE, CS, and HHO, while generating smooth, realizable control signals. These findings confirm MQIO as a potent and reliable optimization tool for advanced robotic control systems, offering significant performance gains for underactuated aerial platforms. Future work will focus on experimental validation and real-time implementation.