Molecular simulation has already become a powerful tool for studying life principles at a molecular level. The past 50-year researches show that molecular simulation has been able to quantitatively characterize the kinetic and thermodynamic properties of complex molecular processes, such as protein folding and conformational changes. In recent years, the application of machine learning algorithms represented by deep learning has further promoted the development of molecular simulation. This work reviews machine learning methods in biomolecular simulation, focusing on the important progress made by machine learning algorithms in improving the accuracy of molecular force fields, the efficiency of molecular simulation conformation sampling, and also the processing of high-dimensional simulation data. The future researches to further overcome the bottleneck of accuracy and efficiency of molecular simulation, expand the scope of molecular simulation, and realize the integration of computational simulation and experimental based on machine learning technique is prospected.