摘要：AbstractBicycles are seen as a viable solution to a number of transportation problems: the use of bicycles alleviate congestion, reduce pollution and generally improve public health. In many congested cities, bicycles are the quickest means of transportation. The widespread use of bicycles, along with the exponential growth of electrically power assisted bicycles, could pose some safety issues related to reckless cycling. This paper presents an inertial measurement based approach to quantify and classify different riding styles. The bicycle velocity, longitudinal acceleration and angular rates are monitored to identify dangerous riding from three different perspective: longitudinal, vertical and lateral dynamics. In this way, it is possible to assign a number assessing the quality of each trip. The approach is developed using data collected from ad-hoc experiments and real-world data.
