Research on the Maximum Regenerative Energy Commutation Control Strategy of a Dual-Mode Synergistic Energy Recovery Pump-Controlled Grinder

Large-inertia pump-controlled grinding machines experience significant energy loss and potential hydraulic shock during frequent high-speed table reciprocation. Traditional control methods often neglect to address efficient energy recovery during the dynamic commutation phase. This study proposes and investigates a dual-mode synergistic energy recovery system that combines motor regeneration and accumulator storage for pump-controlled grinders. The primary focus of this study is on developing a maximum regenerative energy commutation control strategy. A mathematical model of the system was established, and extensive simulations were performed to analyze the energy recovery process under varying load mass, initial velocity, and leakage coefficient conditions. Machine learning models were compared for predicting the peak time of total recovered energy, with a neural network (NN) demonstrating superior accuracy (R 2 ≈ 0.99997). An adaptive commutation strategy was designed, utilizing the NN prediction corrected by a confidence score based on historical and test data ranges, to determine the optimal moment for initiating reverse motion. The strategy was validated using Simulink–Amesim co-simulation and experiments conducted on a 10-ton test bench. The results show that the proposed strategy effectively maximizes energy capture; experiments indicate a 14.3% increase in energy recovery efficiency and a 25% reduction in commutation time compared to a fixed timing approach. The proposed commutation strategy also leads to faster settling to steady-state velocity and smoother operation, while the accumulator demonstrably reduces pressure peaks. This research provides a robust method for enhancing energy efficiency and productivity in pump-controlled grinding applications by improving regenerative braking control through a predictive commutation strategy.