Three-Rigid-Link Manipulator Robot Inverse Dynamic Model Using ANN Based on PID Controller

Authors

M.A. EL-Gohary, Mechanical Engineering Department, Faculty of Engineering, Alexandria University, Egypt. - Mechanical Engineering Department, Faculty of Engineering, Beirut Arab University, Lebanon.Follow
M.S. Elksas, Faculty of Engineering / Mansoura University Egypt,Follow
A. E. Banna, Faculty of computer and information Arab Academy for Science, Technology and Maritime TransportationFollow
S.G. Ahmad, Faculty of Engineering / Mansoura University Egypt,

DOI

10.70259/engJER.2024.851860

Abstract

This work proposes an inverse dynamics model of a three-rigid -link manipulator robot employing a PID controller based on artificial neural networks. In order to attain precise position, neural network control algorithms are designed to handle nonlinear issues including robot manipulator control and uncertainty compensation. An artificial neural network with two layers that uses feedforward training is trained using the back propagation technique. By just fusing the ANN with other traditional control techniques, the suggested controller gives the network more information on the composition and dynamics of the system. The PID controller generates data that is used to train neural network. The simulation's outcome demonstrates that using an ANN in conjunction with a PID controller enhances performance and outperforms a traditional PID.

Recommended Citation

EL-Gohary, M.A.; Elksas, M.S.; Banna, A. E.; and Ahmad, S.G. (2024) "Three-Rigid-Link Manipulator Robot Inverse Dynamic Model Using ANN Based on PID Controller," Journal of Engineering Research: Vol. 8: Iss. 5, Article 10.
DOI: 10.70259/engJER.2024.851860
Available at: https://digitalcommons.aaru.edu.jo/erjeng/vol8/iss5/10