Advanced control of twin rotor multi-input multi-output systems using seagull optimization for linear quadratic regulator tuning

• Дата: 2026
• Автори: Mostefaoui, H., Tahraoui, S., Souaihia, M., Taleb, R., Mostefaoui, M.
• Формат: Стаття
• Мова: Англійська
• Опубліковано: National Technical University "Kharkiv Polytechnic Institute" and Аnatolii Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine 2026
• Теми: seagull optimization algorithm; linear quadratic regulator; twin rotor multi-input multi-output system; parameter tuning; control performance
• Онлайн доступ: http://eie.khpi.edu.ua/article/view/347616
• Назва журналу: Electrical Engineering & Electromechanics

Introduction. During the past decade, advanced control of complex multi-input multi-output (MIMO) systems has been a sustained focus owing to their growing use in aerospace and robotic platforms. The twin rotor MIMO system (TRMS) serves as a helicopter-like benchmark system for testing advanced control techniques. Its nonlinear behavior and significant cross-coupling render it difficult to control using traditional methods. Problem. The TRMS features strong nonlinear dynamics and cross-coupling effects that challenge conventional control methods. Manual tuning of control parameters often results in suboptimal performance and reduced robustness. The goal of this study is to optimize the linear quadratic regulator (LQR) weighting matrices Q and R for the TRMS using the seagull optimization algorithm (SOA) to improve transient performance, minimize overshoot, and accelerate stabilization in both pitch and yaw compared to classical LQR tuning. Methodology. The new approach integrates the SOA with LQR control theory. The SOA determines the best values of Q and R matrices by minimizing a cost function defined by system performance metrics. SOA-optimized LQR is evaluated through simulations and contrasted with the classical LQR under identical conditions. Population size is 50 agents with a maximum of 100 iterations to achieve convergence. Results. Simulation results show that the SOA-optimized LQR has a remarkable improvement in the system’s time response. In comparison to the classical LQR, these results provide a shorter settling time from 7.35 s to 5.34 s (≈28 %), decreases overshoot (≈3 % vs. 30 % open loop), increases damping, and reduces oscillations. The pitch and yaw angle responses across several control schemes clearly demonstrate the superior performance of the proposed optimization technique. Scientific novelty. This work demonstrates, for the first time, the use of SOA for optimal tuning of LQR in a TRMS benchmark. It opens new avenues to enhance the performance of high-order nonlinear systems, pointing toward more accurate and stable control techniques in industrial and aerospace engineering fields. Practical value. The technique provides an efficient method to enhance the functionality of complex nonlinear systems without requiring manual tuning, and it has potential applications in the industrial and aerospace areas. References 38, tables 3, figures 4.