Pressure prediction on a variable-speed pump controlled hydraulic system using structured recurrent neural networks

• Published in: Control engineering practice Vol. 26; pp. 51 - 71
• Main Authors: Kilic, Ergin, Dolen, Melik, Caliskan, Hakan, Bugra Koku, Ahmet, Balkan, Tuna
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-05-2014
• Subjects: Acceptability; Actuators; Computational fluid dynamics; Cylinders; Hydraulic systems; Kalman filtering; Long-term pressure prediction; Networks; Nonlinear system identification; Pressure dynamics; Pump controlled hydraulic system; Pumps; Recurrent neural networks; Structured recurrent neural networks; Pressure dynamics; Kalman filtering; Pump controlled hydraulic system; Long-term pressure prediction; Structured recurrent neural networks; Nonlinear system identification
• ISSN: 0967-0661; 1873-6939
• DOI: 10.1016/j.conengprac.2014.01.008

This paper presents a study to predict the pressures in the cylinder chambers of a variable-speed pump controlled hydraulic system using structured recurrent neural network topologies where the rotational speed of the pumps, the position and the average velocity of the hydraulic actuator are used as their inputs. The paper elaborates the properties of such networks in extended time periods through detailed simulation- and experimental studies where black-box modeling approaches generally fail to yield acceptable performance. As alternative estimation techniques, both linear- and extended Kalman filters are considered in this paper. The estimation properties of the devised network models are comparatively evaluated and their potential application areas are discussed in detail.
•Models to predict pressure in cylinder chambers of a hydraulic system are devised.•Black-box models are found to be insufficient for long-term prediction of pressures.•A structured neural network is proposed to predict pressure dynamics accurately.•Proposed network could be easily adapted to model other similar hydraulic systems.