Distributed physical human machine interaction using intelligent pneumatic cylinders

The increasing use of mechanical devices using compressed air demands for better control approach and faster responds. In this paper, the development and experimental evaluation of an intelligent pneumatic cylinder for distributed physical human machine interaction is proposed. The system applied 36 links of the intelligent pneumatic cylinders to form an Intelligent Chair Tool (ICT) application where each cylinder consists of optical encoder, pressure sensor, valves and a programmable system on chip (PSoC) microcomputer. The cylinder is unique from each other therefore distributed model is essentially implemented. Four extensive control approaches are proposed and experimentally evaluated namely position servo control, force control, compliance control and viscosity control. The control methodology presented contains an inner force loop and an outer position loop implemented using unified control system driven by PWM to an on/off valve. PI controller is used in the force loop to nullify the nonlinearity arising from the compressibility of air. An emulation experiment using mass was also carried out and the results clearly show the ability of the control approaches to verify the future ICT application.