Abstract:Based on the data from traditional household travel survey (HTS) of Mianyang City of Sichuan Province in China in 2010, the approach and process for extracting activity based data from traditional HTS are discussed. A hierarchy selection model system, which consists of mode/destination choice model, time choice model and daily activity pattern choice model, is proposed. Finally, the activity based traffic demand forecasting is achieved by the complete activity based analysis approach proposed with detailed explanation of data processing, model structure, variable selection, and model calibration process.