The invention relates to a traffic trip feature extraction method based on multi-mode public transport data matching. According to the traffic trip feature extraction method based on multi-mode public transport data matching, the concept of a public transport commuting trip chain is introduced, and resident commuting trip feather information such as the commuting trip time, the commuting trip distance and the transfer characteristics is obtained by conducting data pre-processing and matching, public transport trip chain structure extraction, commuting trip behavior judgment, trip origin-destination determination and trip feather information matching according to the multi-source data including ground public transport, rail transport and public bike card swiping data, public transport GPS data, public transport stations, line basic data and the like. According to the traffic trip feature extraction method based on multi-mode public transport data matching, the resident commuting trip condition can be mastered by public transport operating enterprises and government competent departments accurately, the support is provided for optimizing urban public transport lines and making public transport policies, and the traffic trip feature extraction method has great significance in improving commuting trip efficiency of residents.