Embedded Artificial Neural Network on PLCs to Predict Nonlinear System Responses
Conference paper > Full Text · 2020
Authors: Okan DUYMAZLAR, Mustafa ENGİN, DİLŞAD ENGİN
2020 9th Mediterranean Conference on Embedded Computing (MECO), Montenegro, International, Verbal, Full Text, 2020-2020, WoS, vol 1, pp 1-4
DOI: 10.1109/meco49872.2020.9134167
Summary
In this study, an artificial neural network, which runs on a real-time controller to obtain outputs of nonlinear multivariable system is presented. Within the scope of the study, forward kinematics problems of a 6 DOF articulated manipulator has been used for the case study of a nonlinear system. An artificial neural network has been trained for forward kinematics problem and implemented on a PLC to predict the responses of the nonlinear systems on real-time. Also an HMI was used for the observation of system responses and serve as an input device. As a conclusion, an embedded real-time system that successfully works with trained artificial neural network, has been implemented on a common nonlinear case.