Quadratic Programming to Optimize Energy Efficiency of Speed- and Displacement-Variable Pumps

Within the last years, speed-variable pump drives were investigated in numerous applications. In combination with a variable displacement pump, the volume flow and the drive speed can be decoupled. In this paper the resulting degree of freedom will be used to minimize the energy consumption of hydraulic processes by means of a novel model predictive control concept. A dynamic loss model of all drive components will be transformed to a mathematical quadratic optimization problem. The optimum use of the two control variables can achieve energy savings of up to 25% in comparison to known control strategies of speed-variable variable-displacement pumps. Especially in highly dynamic process cycles the proposed optimization guarantees optimum energy efficiency while known approaches become inefficient.