Electro-Hydraulic Position Tracking using NARMA Neural Controller
Ndilkoula Diontar1, Christopher A. Otieno2, Stanley I. Kamau3

1Ndilkoula Diontar, Pan Africa University Institute for Basic Science Technology and Innovation. Mechatronics Engineering Department
2Christopher Otieno Adika, Multi Media University, Engineering Department
3Stanley. I. Kamau, Jomo Kenyatta University of Agriculture and Technology. Electrical Engineering Department.

Manuscript received on April 15, 2020. | Revised Manuscript received on April 20, 2020. | Manuscript published on May 30, 2020. | PP: 48-52 | Volume-9 Issue-1, May 2020. | Retrieval Number: F8544038620/2020©BEIESP | DOI: 10.35940/ijrte.F8544.059120
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: This paper presents an electro-hydraulic position tracking which is one of the most used applications in industries such as automobile, aeronautic, robotic, computer numeric control etc. It is used widely in industrial application due to its higher force and torque generation, smooth response characteristic and good positioning capabilities. In order to design a controller to track the position of the system, a mathematical model of the system was first developed. From this model a nonlinear state space of the system was found and simulated in open loop in MATLAB/SIMULINK. After developing the mathematical model, Nonlinear Auto Regression Moving Average (NARMA) Neural Controller based which is able to cancel out the nonlinearity of the electro hydraulic by transforming the nonlinear system dynamic into linear system dynamic was designed to control the electro hydraulic plant. In the neural controller design process, first a neural network was trained offline and then the trained neural network was reconfigured as a controller to track the reference. After the controller eliminates the nonlinearity and dynamic of the system, the input output relation become a simple implicit relation and the output of the plant was able to track the reference. In order to evaluate the performances of the designed controller, a Proportional Integral (PI) controller was tuned and its response was compared with the one of NARMA neural controller. Results showed that NARMA neural controller based presents a better overshoot and settling compare to proportional integral controller.
Keywords: Electro-hydraulic, Neural Network, NARMA, Proportional Integral controller.
Scope of the Article: Sensor Networks, Actuators for Internet of Things.