A travel mode identification method based on GPS trajectory data

Abstract

The invention discloses a travel mode identification method based on GPS trajectory data. For travel modes with missing or incomplete GPS signals, a rule-based subway identification method is adopted; for other travel modes with normal GPS signal records, a random forest classifier is used , combined with GPS trajectory data for modeling and verification. Before mode identification, the feature selection process screens out 4 types of feature parameters related to mode identification, which can accurately identify all real travel modes. This method is not limited by GPS data, does not rely on other data sources, has high versatility, simple calculation, and accurate and reliable recognition results. It is suitable for travel mode recognition based on GPS trajectory data and is conducive to promoting large-scale promotion of travel surveys based on smartphones. .

Description

technical field [0001] The invention relates to computer identification technology, in particular to a travel mode identification method based on GPS trajectory data. Background technique [0002] With the popularization of smart phones and the development of GPS positioning technology, more and more attention has been paid to the GPS travel survey based on smart phones. This method can make up for the deficiencies in many traditional residents' travel surveys. Investigators have less burden and so on. [0003] The collected GPS trajectory data needs to be processed by a specific algorithm to obtain the information required for travel research. The identification of travel mode is the most critical step. At present, many researchers use data such as GPS, GIS, GIS+GPS, and accelerometers to identify travel modes, and the identification methods used are mainly based on rules and calculation methods. However, because GIS data is relatively difficult to obtain in China, and w...

AI Technical Summary

Problems solved by technology

However, because GIS data is relatively difficult to obtain in China, and when using GIS data and GPS trajectory matching to identify travel modes such as bus mode identification, the calculation is complicated, the efficiency is low, and the identification effect is not good

The rule-based method is relatively dependent on the data. Once the data source changes, the obtained rules are no longer applicable, and the method has poor versatility. Relevant feature parameters are applied to the same recognition model and overfitting occurs, resulting in a decrease in recognition accuracy

Images