Photovoltaic (PV) power generation can convert sunlight into electrical energy. There is an ever developing rapidly industry under the background of carbon peak and carbon neutrality. However, dust accumulation on the surface of PV modules can pose a serious threat to the efficiency of power generation. A hydrophobic coating with low surface energy can be expected to form on the surface, in order to reduce the dust accumulation for better self-cleaning of the PV module. In this study, a mechanical model of the adhesion between the surface of PV modules and dust particles was established to clarify the effect of hydrophobic coating surface properties on the self-cleaning ability of PV modules. The particle contact mechanics was selected to determine the adhesion force of dust particles, particularly with the elastic modulus, Poisson's ratio, and friction coefficient of surface materials. The simulation showed that the surface self-cleaning performance of PV modules was closely related to the elastic modulus and friction coefficient of surface materials, as well as the particle size of dust particles. There was also a strong relationship among the self-cleaning performance of PV modules, dust particle size, and surface performance. The surface elastic modulus played a major role in the self-cleaning performance of PV modules in the range of small particle size dust below 200 μm. With a larger particle size than 200 μm, the friction coefficient was dominated in the self-cleaning performance of PV modules. The self-cleaning of the PV modules varied in the different hydrophobic coatings, and the performance parameters of the PV module surface. The self-cleaning dust particles above 100 μm performed the best using the hydrophobic coating with the friction coefficient of 0.1-0.15 and elastic modulus of 0-100 MPa, whereas the self-cleaning dust particles above 500 μm were required for the hydrophobic coating with the friction coefficient of 0.1-0.15 and elastic modulus of 100-73 000 MPa. The hydrophobic coating with a friction coefficient of 0.1-0.45 and elastic modulus of 100-73 000 MPa was suitable for the PV modules to self-clean dust particles above 1 000 μm. The general soil particle size was divided into fine (100-250 μm), medium (250-500 μm), coarse (500-1 000 μm), very coarse sand (1 000-2 000 μm), and gravel (>2 000 μm). The optimal surface properties of hydrophobic coatings were obtained for the self-cleaning in the different dust particle sizes range. Specifically, the larger the particle size of dust was, the wider the range of elastic modulus and friction coefficient of the hydrophobic coating were. Consequently, the better surface properties of hydrophobic coatings were achieved in the elastic modulus of 0-100 MPa and the friction coefficient of 0.1-0.15, in order to clean the fine sand. The elastic modulus of 100-73 000 MPa and the friction coefficient of 0.1-0.6 were to clean the gravel. The PV power stations can be expected to select the hydrophobic coating for the self-cleaning requirements. The local dust particles can be optimized to reduce the loss of power generation efficiency for the cost saving of PV power stations. For example, once the dust of PV power plants was mainly medium sand (250-500 μm) in Northwest China, the surface properties of hydrophobic coatings can be the elastic modulus of 0-2 700 MPa and the friction coefficient of 0.1.