The study focuses on the optimization of liquid hydrogen (LH2) refueling stations by utilizing a gas-liquid hydrogen mixed pre-cooling strategy to reduce energy consumption. Traditional stations necessitate energy-intensive refrigeration systems for pre-cooling, whereas LH2 stations can use the cold energy released during hydrogen vaporization, leading to zero-energy pre-cooling. The research establishes a thermodynamic model using MATLAB to evaluate the impact of different refueling strategies on station performance and energy efficiency. An Artificial Neural Network optimized by a genetic algorithm is developed to predict state of charge and energy consumption under varying conditions. By selecting optimal initial parameters through an improved genetic algorithm, the study achieved a significant reduction in pump energy consumption, providing valuable insights for developing low-energy refueling strategies in LH2 stations. The article discusses the advantages of LH2 refueling stations over gaseous stations, the importance of pre-cooling hydrogen to comply with industry standards, and the challenges and solutions related to compression equipment in LHRSs. It also highlights the significance of the gas-liquid hydrogen mixing method for pre-cooling and the role of artificial neural networks in simulating and optimizing hydrogen refueling stations.