Motion control design for hydraulic drives remains to be a complicated task, and has not evolved on a level with electrical drives. When considering specifically motion control of hydraulic drives, the industry still prefers conventional linear control structures, often combined with feed forward control and possibly linear active damping functionalities. However difficulties often arise due to the inherent and strong nonlinear nature of hydraulic drives, with the more dominant being nonlinear valve flow- and oil stiffness characteristics, and furthermore the volume expansion/retraction when considering cylinder drives. A widely used approach with electrical drives is state controlor cascade control, that may by successfully applied to manipulate the drive dynamics in order to achieve high bandwidths etc., due to the nearly constant parameter-nature of such drives. Such properties are however, unfortunately not present in valveoperated hydraulic drives.
This paper considers a cascade control approach for hydraulic valve-cylinder drives motivated by the fact that this may be applied to successfully suppress nonlinearities. The drive is pre-compensated utilizing a pressure updated inverse valve flow relation, ideally eliminating the system gain variation, and the linear model equations for the pre-compensated system is used for the cascade control design. The cascade design does not utilize e.g. bode plots, root loci etc., and is based on an analytic approach, emphasizing the exact influence of each control parameter, resulting in an easily comprehensible control structure. Two versions of this cascade control approach is presented, with the first utilizing pressure-, piston velocity- and piston position feedback, and the second utilizing only pressure- and piston position feedback. The latter may be especially interesting in an industrial context, as this does not use the velocity feedback, which is rarely available here. Furthermore, the position control loop is designed analytically to guarantee a user defined gain margin. The proposed control design approach is verified through simulations, and results demonstrate the announced properties.
Robot Structure and Motion Control Design Based on UG and Proteus
