The performance level of traffic control systems highly relies on accurate short-term network-wide traffic prediction.Signal processing techniques have been widely integrated with deep learning algorithms for this pur...The performance level of traffic control systems highly relies on accurate short-term network-wide traffic prediction.Signal processing techniques have been widely integrated with deep learning algorithms for this purpose,but no study has focused on how the parameter of wavelet decomposition order(level)affects the prediction robustness given the training data limitations.This study proposes a hybrid framework for short-term network-wide traffic state prediction,which applies multi-layer perceptron(MLP)neural networks to implicitly capture the traffic network co-movement patterns.Then,the framework is complemented by a seasonal auto-regressive integrated moving average(SARIMA)model,extracting the location-specific features,including the local seasonality and stochastic disturbances.Besides,the hybrid framework is used to explore the association between the traffic prediction accuracy and wavelet decomposition order using an order-adaptive discrete haar wavelet transform(DHWT).The proposed method was validated over four open-access datasets with different training data characteristics in Paris and Madrid urban areas.The results indicated that the hybrid framework significantly improved the predictive accuracy of the benchmark deep learning algorithms.The forecasts for the low-resolution dataset experienced a noticeable improvement once higher wavelet orders were applied.The statistical analysis revealed the moderating effects of the training sample size and data spatial resolution on the link between the wavelet decomposition orders and the model predictive performance.Nevertheless,a combination of pre-processing order of two,and post-processing order of three led to satisfactory results in the cases where sufficient historical data were available.展开更多
Social Media have increasingly provided data about the movement of people in cities making them useful in understanding the daily life of people in different geographies.Particularly useful for travel analysis is when...Social Media have increasingly provided data about the movement of people in cities making them useful in understanding the daily life of people in different geographies.Particularly useful for travel analysis is when Social Media users allow(voluntarily or not)tracing their movement using geotagged information of their communication with these online platforms.In this paper we use geotagged tweets from 10 cities in the European Union and United States of America to extract spatiotemporal patterns,study differences and commonalities among these cities,and explore the nature of user location recurrence.The analysis here shows the distinction between residents and tourists is fundamental for the development of city-wide models.Identification of repeated rates of location(recurrence)can be used to define activity spaces.Differences and similarities across different geographies emerge from this analysis in terms of local distributions but also in terms of the worldwide reach among the cities explored here.The comparison of the temporal signature between geotagged and non-geotagged tweets also shows similar temporal distributions that capture in essence city rhythms of tweets and activity spaces.展开更多
文摘The performance level of traffic control systems highly relies on accurate short-term network-wide traffic prediction.Signal processing techniques have been widely integrated with deep learning algorithms for this purpose,but no study has focused on how the parameter of wavelet decomposition order(level)affects the prediction robustness given the training data limitations.This study proposes a hybrid framework for short-term network-wide traffic state prediction,which applies multi-layer perceptron(MLP)neural networks to implicitly capture the traffic network co-movement patterns.Then,the framework is complemented by a seasonal auto-regressive integrated moving average(SARIMA)model,extracting the location-specific features,including the local seasonality and stochastic disturbances.Besides,the hybrid framework is used to explore the association between the traffic prediction accuracy and wavelet decomposition order using an order-adaptive discrete haar wavelet transform(DHWT).The proposed method was validated over four open-access datasets with different training data characteristics in Paris and Madrid urban areas.The results indicated that the hybrid framework significantly improved the predictive accuracy of the benchmark deep learning algorithms.The forecasts for the low-resolution dataset experienced a noticeable improvement once higher wavelet orders were applied.The statistical analysis revealed the moderating effects of the training sample size and data spatial resolution on the link between the wavelet decomposition orders and the model predictive performance.Nevertheless,a combination of pre-processing order of two,and post-processing order of three led to satisfactory results in the cases where sufficient historical data were available.
基金partially funded by the DAAD Project(No.57474280)Verkehr-SuTra:Technologies for Sustainable Transportation,within the Programme:A New Passage to India—Deutsch-Indische Hochschulkooperationen ab 2019the German Federal Ministry of Education and Research,Bundesministerium für Bildung und Forschung(BMBF),project FuturTrans:Indo-German Collaborative Research Center on Intelligent Transportation Systemsby the European Union's Horizon 2020 research and innovation programme under grant agreement No.815069(project MOMENTUM(Modelling Emerging Transport Solutions for Urban Mobility)).
文摘Social Media have increasingly provided data about the movement of people in cities making them useful in understanding the daily life of people in different geographies.Particularly useful for travel analysis is when Social Media users allow(voluntarily or not)tracing their movement using geotagged information of their communication with these online platforms.In this paper we use geotagged tweets from 10 cities in the European Union and United States of America to extract spatiotemporal patterns,study differences and commonalities among these cities,and explore the nature of user location recurrence.The analysis here shows the distinction between residents and tourists is fundamental for the development of city-wide models.Identification of repeated rates of location(recurrence)can be used to define activity spaces.Differences and similarities across different geographies emerge from this analysis in terms of local distributions but also in terms of the worldwide reach among the cities explored here.The comparison of the temporal signature between geotagged and non-geotagged tweets also shows similar temporal distributions that capture in essence city rhythms of tweets and activity spaces.
