In recent years, the emergence and development of single-molecule tracking technology has provided a new way to study the biological process in a cell membrane. However, the heterogeneity of biomolecular motions in cell membranes makes it difficult to distinguish and separate different molecular motion patterns from a large number of molecular trajectories. In this work, taking the system of interaction between melittin and a single-component supported phospholipid bilayer (SLB) membrane as an example, a type of data analysis method is developed through using the frequency distribution of standard deviations of single molecular motion displacements to distinguish and separate lipid molecules in different motion modes. It provides higher accuracy and more quantitative information than traditional analysis methods based on the frequency distribution of displacement or radius of gyration. By using this method, we successfully separate the two moving states of lipids in the SLB, and find that such a distribution is partly consistent with the location distribution of lipids in the upper leaflet and lower leaflet. Under the influence of surface adsorption or transmembrane poration of melittin at different concentrations, the movements of these two types of lipids are disturbed in different ways. In this work, a data analysis method for the separation of molecular motion patterns is developed according to the complex heterogeneity of molecular motions in a phospholipid membrane system. The different effects on the upper leaflet and lower leaflet of a lipid membrane due to melittin actions at different stages are obtained. The method developed in this work will be of great help in studying the dynamic processes of biological systems by using single-molecule tracking technology.