Energy Saving Strategy of the Variable-Speed Variable-Displacement Pump Unit Based on Neural Network

Abstract The hydraulic system is widely used in manufacturing fields, and the hydraulic drive unit is one of the energy intensive components in the hydraulic system. For reducing the energy consumption, a variable-speed variable-displacement pump unit (SVVDP) was investigated. The optimum energy efficiency of the unit can be realized by regulating the motor rotating speed and the pump displacement simultaneously. However, it is difficult to find the optimal rotating speed and displacement for each working condition. In this paper, the problem is solved by developing the control strategy using a backpropagation neural network, which is utilized to calculate the speed and displacement based on the measured pressure and flow rate of the hydraulic system. Results indicate that the proposed strategy can reliably and automatically lower energy consumption of the SVVDP under various conditions. The proposed control strategy contributes to lowing energy consumption of various types of hydraulic equipment and construction machinery.