Time series clustering model based on DTW for classifying car parks

14Citations
Citations of this article
25Readers
Mendeley users who have this article in their library.

Abstract

An increasing number of automobiles have led to a serious shortage of parking spaces and a serious imbalance of parking supply and demand. The best way to solve these problems is to achieve the reasonable planning and classify management of car parks, guide the intelligent parking, and then promote its marketization and industrialization. Therefore, we aim to adopt clustering method to classify car parks. Owing to the time series characteristics of car park data, a time series clustering framework, including preprocessing, distance measurement, clustering and evaluation, is first developed for classifying car parks. Then, in view of the randomness of existing clustering models, a new time series clustering model based on dynamic time warping (DTW) is proposed, which contains distance radius calculation, obtaining density of the neighbor area, k centers initialization, and clustering. Finally, some UCR datasets and data of 27 car parks are employed to evaluate the performance of the models and results show that the proposed model performs obviously better results than those clustering models based on Euclidean distance (ED) and traditional clustering models based on DTW.

Cite

CITATION STYLE

APA

Li, T., Wu, X., & Zhang, J. (2020). Time series clustering model based on DTW for classifying car parks. Algorithms, 13(3). https://doi.org/10.3390/a13030057

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free