Identifying travel modes from massive information is critical for travel demand modeling and transportation planning. Along with widespread application of the highly accurate position devices, the need of cost-effective approaches is more urgent than ever before. In the field of movement ecology, it has been proved by experiments that Hidden Markov Model (HMM) is much more efficient than traditional statistical classification method, such as Bayesian methods. Accordingly, in this paper, the Hidden Markov Model is adopted to predict travel modes based on raw GPS trajectories, where the modes are labeled as walk, bike, bus, and driving. Besides, this study also endeavors to examine the inherent statistical characteristics of trajectories in different modes, and identify the relationship between residents’ choice of travel mode and the traffic flow structure along the route. The data analysis results are believed to gain a deeper insight into the impact of distance and route on residents’ choice of travel mode, and it is expected to provide ideas and references for the transportation planning of cities with completed transportation infrastructure construction.
Inferring Travel Modes from Trajectory Data Based on Hidden Markov Model
International Conference on Transportation and Development 2020 ; 2020 ; Seattle, Washington (Conference Cancelled)
31.08.2020
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Inferring Travel Modes from Cellular Signaling Data Based on the Gated Recurrent Unit Neural Network
DOAJ | 2023
|Inferring Traffic Signal Phases From Turning Movement Counters Using Hidden Markov Models
Online Contents | 2015
|Modelling changes in travel behaviour mechanisms through a high-order hidden Markov model
Taylor & Francis Verlag | 2024
|