Micro-circulation bus is a kind of public transportation with the advantages of flexible operation organization and high turnover efficiency. It is one of the important solutions for the“first and last mile” problem of citizens. In response to this problem, the bike-sharing system has risen rapidly in recent years, resulting in massive data with numerous short-distance travel demands. Based on the bike-sharing data and bus stations data in Chengdu, this paper analyzes individual travel characteristics and identify the commuting patterns by using clustering analysis and data outlier analysis methods. Combining the above results with the operational characteristics of micro-circulation bus，a genetic algorithm is adopted to plan micro-circulation bus routes. With the support of data, more reasonable micro-circulation bus routes can be automatically generated and diversified development of public transportation can be promoted. Keywords：bike-sharing, big data, travel characteristics analysis and micro-circulation bus route planning

CICTP