Home > Published Issues > 2021 > Volume 10, No. 1, January 2021 >

Neural Control for Image Stabilisation Using a Reference Model

Gani Balbayev 1, Aigerim Mussina 1, Algazy Zhauyt 1, Beibit Shingissov 1, and Marina Kalekeyeva 2
1. Almaty University of Power Engineering and Telecommunications, Almaty 050013, Kazakhstan
2. Academy of Civil Aviation, Almaty 050049, Kazakhstan

Abstract—In this paper, the mathematical interpretation of VOR is derived and a VOR adaptation method is newly proposed. This study of cerebellar inspired adaptive control, which is augmented by the reference model, provides a potential general solution for robotic control. If the reference model is not used, the plant with more poles than zeroes finds it difficult to give a response immediately. The reference model determines the behaviour of the robot. The response of it at high frequency with small-time constant shows the roll-off, low frequency signals not affected. To verify the proposed method of neural control for image stabilisation using a reference model, first, it is briefly introduced and then it is utilized for Matlab and Simulink simulation and experiments. The described cerebellar algorithm has the potential to provide a modular controller for soft robots. Also, it can be applied to generic control tasks, as the force control, or impedance response, or task calibration.

Index Terms—vestibulo-ocular reflex, cerebellum, model reference adaptive control, robotics

Cite: Gani Balbayev, Aigerim Mussina, Algazy Zhauyt, Beibit Shingissov, and Marina Kalekeyeva, "Neural Control for Image Stabilisation Using a Reference Model," International Journal of Mechanical Engineering and Robotics Research, Vol. 10, No. 1, pp. 17-21, January 2021. DOI: 10.18178/ijmerr.10.1.17-21