Optimal cylinder commutation in digital hydraulic pumps and motors for pulsation minimization

Digital Hydraulic Pumps and Motors are the digital equivalent to conventional analog hydrostatic displacement machines. They use switching on-off valves instead of continuously variable valves, yielding mechanical robustness, fast dynamics, power efficiency and price competitiveness. Because of the discrete inputs, these benefits however come at the expense of a challenging control design. State of the art control schemes for commutation of the hydraulic cylinders introduce undesirable pulsations which inhibit a proper machine function and can inflict damage. This work shows a new control approach for mitigating pulsations based on model predictive optimization. Not only are the pulsation amplitudes controlled, but the presented methodology also allows for an explicit consideration of frequency content in the produced output. The discrete characteristics of the digital hydraulic pump/motor are modelled by a hybrid dynamical system model which is translated into a system of difference equations subject to linear inequalities with mixed-integer input variables. This Mixed Logical Dynamical system model is used as a prediction model in an optimization to reduce pulsation amplitudes and set desired frequencies, effectively solving a Mixed-Integer Quadratic Program. The result are switching sequences that yield an output volume flow or torque with desired properties.