Uluslararası Mühendislik Sempozyumu, İzmir, Türkiye, 5 - 13 Aralık 2020, ss.1-2
Due to the high power and low weight-to-power ratio, electrohydraulic systems are an indispensable part of the industry. The introduction of efficient control methods plays a vital role in the effective utilization of electrohydraulic systems as well as improving their structural properties. For that purpose, the PID controller emerges as a widely adopted method for electrohydraulic systems. On the other hand, the poor control performance of the PID controller against system uncertainties and nonlinear effects adversely affects the success of the electrohydraulic systems. Nowadays, control methods that are robust to disturbance and easy-to-design are introduced. The fractional-order PID controller stands out among the methods presented. The fractional-order PID controller offers features that meet the industry’s requirements: (i) A higher number of adjustable coefficients than a traditional PID controller, (ii) resistant to system uncertainties, and (iii) the simplistic design process. Thus, an electrohydraulic test rig is employed to compare the efficiency of the conventional PID and the fractional-order PID in terms of their control performance. The electrohydraulic system consists of a differential hydraulic piston driven by a proportional valve and a fixed speed - constant displacement pump. In this study, the position control of the differential hydraulic piston is carried out experimentally. The obtained electrohydraulic system model, which is utilized for the controller design, solely relies on the catalog data to examine the controllers’ effectiveness against system uncertainties. The obtained model does not include (i) internal leaks, (ii) piston friction, (iii) proportional valve hysteresis and spool position uncertainty, (iv) flow fluctuations due to the pump drive system, and (v) temperature-dependent characteristics of the hydraulic fluid. The particle swarm optimization approach is employed to determine the controller coefficients through the obtained system model. Also, step and sine wave input signals are applied as position reference, while the integral of absolute error (ISE) is adopted as a performance criterion throughout the study. As a consequence, it has been revealed that the fractional-order PID controller performs better than the PID controller against the uncertainties due to the unmodeled system dynamics.