This paper focuses on the passive ventilation system modeling for Hangzhou Science and Technology Pavilion of Green Architecture. The input and output variables of the model are determined according to the passive ventilation principles and the correlation analysis among the input and output data under various climatic conditions. The passive ventilation model with multiple inputs is built. Its consistency is validated theoretically and experimentally. Office temperature of the next day is predicted depending on the weather forecast parameters obtained from the Internet. Thus, precise time to start airconditioning systems can be determined for energy efficiency.