Traffic Status Evaluation Based on Possibilistic Fuzzy C-Means Clustering Algorithm

摘要

Intelligent Transportation System (ITS) is a principal part of smart city, and traffic status evaluation is an essential role in intelligent transportation management. Nowadays, a number of models and algorithms based on traffic flow theories and machine learning were applied to evaluate the traffic status. However, the evaluation results of the two types of methods either take high computational cost or will be easily affected by noise. To overcome these drawbacks, a new traffic status evaluation model based on possibilistic fuzzy c-means (PFCM) is proposed in this paper. The dataset from Caltrans Performance Measurement System (PeMS) is used in experiments. The PFCM algorithm is compared with fuzzy c-means (FCM) algorithm to validate its accuracy and anti-noise capabilities. Experiment results show that the proposed model can evaluate traffic status more efficiently.