Visual sensing based adaptive sliding mode control of position tracking in electrohydraulic systems

This study focuses on the issue of position tracking of a budgeted electrohydraulic system with the aid of marker based visual sensing of the actuator rod position using a single camera. The system comprises of a single-rod cylinder having non-linear friction characteristics and industry grade proportional valve possessing a large-deadband. A controller has been developed based on an input linearized system model involving the combination of a sliding mode and a deadband controller. An adaptive strategy for the controller parameters relying on a Lyapunov function has been introduced to handle the nonlinearities and uncertainties involved in the system. The controller demonstrated satisfactory performance for the realtime experiments conducted on the laboratory scale system.