Modeling and bumpless learning control of position and pressure for an electropneumatic actuator

This paper presents the modeling, simulation and control of a pneumatic cylinder actuator driven by a high-performance pneumatic servovalve. The modeling is done via a combination of first principles and empirically identified system characteristics. A stick-slip friction model is incorporated as a necessary component of the overall system description. The model is verified against experimental data and shown to be quite accurate over a range of operating conditions. Two separate Iterative Learning Controllers (ILC's) are applied to the system to improve the overall sequential tracking of position and pressure reference inputs. In addition, a bumpless transfer technique is used to transition between the position and pressure control loops in a smooth manner. The profiles chosen are similar to those found in certain manufacturing processes such as blow molding. It is show via simulation and experiment that the ILC can provide an improvement in the system tracking capabilities and the bumpless transfer can be used very effectively to minimize transients in the changeover from position to pressure.