Comparative Analysis of Robust, Adaptive, and Neural Network Control of a PUMA 560 Robot Manipulator under Payload Variation

Öz

Robot manipulators are required to operate under uncertain and time-varying dynamic conditions where payload variations, modeling uncertainties, and external disturbances may significantly affect trajectory tracking performance. This study presents a comparative analysis of robust, adaptive, and neural network-based control strategies for a six-degree-of-freedom (6-DOF) PUMA 560 robot manipulator subjected to time-varying payload disturbances. All controllers were implemented in MATLAB/Simulink and evaluated under identical joint-space reference trajectories and disturbance conditions. During the simulation, the manipulator performs a pick-and-place task in which a 1 kg object is grasped by the end-effector at t = 3 s and released at t = 6 s. Therefore, the payload affects robot dynamics only between 3 s and 6 s. This scenario creates a more realistic time-varying payload condition and allows the disturbance rejection and adaptation capabilities of the controllers to be evaluated under a practical robotic manipulation task. Among the uncertainty-tolerant controllers, the neural network controller achieved the lowest overall RMS tracking error while preserving smooth and bounded torque responses, indicating strong nonlinear approximation and online learning capability. Robust control provided reliable disturbance rejection and stable tracking performance under modeling uncertainty, whereas adaptive control maintained stability through online parameter estimation despite a transient adaptation period. Overall, the results indicate that neural network-based control provides a highly promising intelligent control framework for robotic manipulators operating under nonlinear and uncertain dynamic conditions.

Anahtar Kelimeler

• Robust control
• adaptive control
• neural network control
• PUMA 560
• robot manipulators
• payload variation

Etik Beyan

The authors declare that there is no ethical issue.

Yük Değişimi Altında PUMA 560 Robot Manipülatörünün Sağlam, Uyarlanabilir ve Sinir Ağı Kontrolünün Karşılaştırmalı Analizi

Öz

Robot manipülatörlerinin, yük varyasyonları, modelleme belirsizlikleri ve dış bozulmaların yörünge izleme performansını önemli ölçüde etkileyebildiği belirsiz ve zamana bağlı dinamik koşullar altında çalışması gerekmektedir. Bu çalışma, zamana bağlı yük bozulmalarına maruz kalan altı serbestlik dereceli (6-DOF) bir PUMA 560 robot manipülatörü için sağlam, uyarlanabilir ve sinir ağı tabanlı kontrol stratejilerinin karşılaştırmalı bir analizini sunmaktadır. Tüm kontrolörler MATLAB/Simulink'te uygulandı ve aynı eklem uzayı referans yörüngeleri ve bozulma koşulları altında değerlendirildi. Simülasyon sırasında, manipülatör, t = 3 s'de uç efektör tarafından 1 kg'lık bir nesnenin kavranıp t = 6 s'de bırakıldığı bir alma ve yerleştirme görevi gerçekleştirir. Bu nedenle, yük robot dinamiğini yalnızca 3 s ile 6 s arasında etkiler. Bu senaryo, daha gerçekçi bir zamana bağlı yük koşulu oluşturur ve kontrolörlerin bozulma reddi ve uyarlama yeteneklerinin pratik bir robotik manipülasyon görevi altında değerlendirilmesine olanak tanır. Belirsizliğe dayanıklı kontrolcüler arasında, sinir ağı kontrolcüsü, düzgün ve sınırlı tork tepkilerini korurken en düşük genel RMS izleme hatasını elde etti; bu da güçlü doğrusal olmayan yaklaştırma ve çevrimiçi öğrenme yeteneğini gösteriyor. Sağlam kontrol, modelleme belirsizliği altında güvenilir bozucu etki reddi ve istikrarlı izleme performansı sağlarken, uyarlanabilir kontrol, geçici bir adaptasyon dönemine rağmen çevrimiçi parametre tahmini yoluyla kararlılığı korudu. Genel olarak, sonuçlar, sinir ağı tabanlı kontrolün, doğrusal olmayan ve belirsiz dinamik koşullar altında çalışan robotik manipülatörler için oldukça umut vadeden akıllı bir kontrol çerçevesi sağladığını göstermektedir.

Anahtar Kelimeler

• Gürbüz kontrol
• uyarlanabilir kontrol
• sinir ağı kontrol
• PUMA 560
• robot kolu
• yük değişimi

Etik Beyan

Yazarlar herhangi bir etik sorun olmadığını beyan ederler.

Kaynakça

1. J. J. Craig, Introduction to Robotics: Mechanics and Control, 3rd ed. Upper Saddle River, NJ, USA: Pearson Education, 2005.
2. F. L. Lewis, D. M. Dawson, and C. T. Abdallah, Robot Manipulator Control: Theory and Practice, 2nd ed. New York, NY, USA: Marcel Dekker, 2004.
3. K. M. Doğan, E. Tatlıcıoğlu, E. Zergeroğlu, and K. Çetin, “Observer Based Output Feedback Repetitive Learning Control of Robotic Manipulators,” IET Control Theory & Applications, vol. 18, no. 15, 2024.
4. H. Tirsi and K. Çetin, “Model-Based Nonlinear Control Applications with Kinematic and Dynamic Analysis of 4-DoF SCARA Robot Manipulator,” Journal of Scientific Reports-A, 2025.
5. E. Sönmez and K. Çetin, “Adaptive Tracking Control for Dynamic Model Uncertainties of a 6-DOF Industrial Staubli Robot Manipulator Carrying Variable Payloads,” International Journal of Research and Review, vol. 12, no. 9, 2025.
6. K. Çetin, E. Tatlıcıoğlu, and E. Zergeroğlu, “On Operational Space Tracking Control of Robotic Manipulators With Uncertain Dynamic and Kinematic Terms,” Journal of Dynamic Systems, Measurement, and Control.
7. K. M. Doğan, E. Zergeroğlu, E. Tatlıcıoğlu, and K. Çetin, “Learning Control of Robot Manipulators in Task Space,” Asian Journal of Control, vol. 20, 2018.
8. K. Çetin, C. S. Zapico, H. Tuğal, Y. Petillot, M. Dunnigan, and M. S. Erden, “Application of Adaptive and Switching Control for Contact Maintenance of a Robotic Vehicle-Manipulator System for Underwater Asset Inspection,” Frontiers in Robotics and AI, vol. 8, 2021.

9. M. Deniz, “A Model-Free Continuous Velocity Observer Formulation With Self-Tuning for Mechatronic Systems,” Journal of Dynamic Systems, Measurement, and Control, vol. 140, 2018.
10. Y. Tong, J. Liu, H. Zhou, Z. Ju, and X. Zhang, “Adaptive Tracking Control of Robotic Manipulators With Unknown Kinematics and Uncertain Dynamics,” IEEE Transactions on Automatic Control.
11. W. He, B. Huang, Y. Dong, Z. Li, and C. Y. Su, “Adaptive Neural Network Control for Robotic Manipulators With Unknown Deadzone,” IEEE Transactions on Cybernetics, vol. 48, no. 9.
12. P. V. Cuong and W. Y. Nan, “Adaptive Trajectory Tracking Neural Network Control With Robust Compensator for Robot Manipulators,” Neural Computing and Applications, vol. 27, pp. 525–536, 2016.
13. A. K. Dwivedi, U. Singh, and A. Agarwal, “Learning Based Motion Planning of a 6-DOF Manipulator by Using a Neural Network,” Proc. IEEE Applied Sensing Conference, 2026.
14. “Generalization of Neural Network for Manipulator Inverse Dynamics Model Learning,” Applied Intelligence, vol. 55, Art. no. 669, 2025.
15. T. K. Mohammed and C. M. Abdissa, “Robust Model Reference Adaptive Controller for 3-DOF Planar Manipulator,” Scientific Reports, vol. 16, Art. no. 4889, 2026.
16. M. Salman, Z. Niu, R. Singh, L. Kshetrimayum, and I. Hussain, “Robust Control of a Compliant Manipulator With Reduced Dynamics and Sliding Perturbation Observer,” Scientific Reports, vol. 15, Art. no. 8934, 2025.
17. K. Guo, H. Jiang, J. Wang, H. Qin, and P. Zhang, “Adaptive-Based Non-Singular Terminal Sliding Mode Control of a Three-DOF Spatial Robotic Manipulator,” Journal of Field Robotics, vol. 43, no. 1, 2026.
18. K. B. Gaufan, M. B. Aremu, N. M. Alyazidi, and A. Nasir, “Robust Fractional-Order Sliding Mode Control for Robotic Manipulator System With Time-Varying Disturbances,” Franklin Open, vol. 12, Art. no. 100287, 2025.
19. O. Karahan and H. Karci, “Design of Robust Fractional Order Fuzzy PID Sliding Mode Controller Based on Hybrid Swarm Intelligence Algorithm for a 6-DOF Robotic Manipulator,” Robotica, 2025.
20. X. F. Liu, K. H. Wang, J. H. Liu, and G. P. Cai, “Kinematic Recurrent Neural Network-Based Hybrid Calibration for Serial Manipulators,” IEEE Transactions on Instrumentation and Measurement, vol. 75, 2026.
21. J. Pacheco, D. Cortés-Vega, and H. Alazki, “Speed Sensorless Motion Control Scheme for a Robotic Manipulator Under External Forces and Payload Changes,” Actuators, vol. 14, no. 5, Art. no. 209, 2025.
22. W. F. W. Tarmizi, F. Samsuri, and M. L. M. Zain, “Optimization of PID Controller for Robotic Arm Tracking Using Particle Swarm Optimization and Deep Reinforcement Learning,” Proc. IEEE International Conference, 2025.
23. S. Samo et al., “Zero-Shot Payload Adaptation for Robotic Manipulators via Domain-Randomized Deep Reinforcement Learning,” Mehran University Research Journal of Engineering and Technology, vol. 45, no. 2, 2026.
24. S. K. Jha et al., “Non-linear Fractional Order Fuzzy PD Plus I Controller for Trajectory Optimization of 6-DOF Modified Puma-560 Robotic Arm,” Proceedings of Congress on Control, Robotics, and Mechatronics (CRM), pp. 415–431, 2023.