Understanding how congestion at one location can cause ripples throughout large-scale transportation network is vital for transportation researchers and practitioners to pinpoint traffic bottlenecks ...for congestion mitigation. Traditional studies rely on either mathematical equations or simulation techniques to model traffic congestion dynamics. However, most of the approaches have limitations, largely due to unrealistic assumptions and cumbersome parameter calibration process. With the development of Intelligent Transportation Systems (ITS) and Internet of Things (IoT), transportation data become more and more ubiquitous. This triggers a series of data-driven research to investigate transportation phenomena. Among them, deep learning theory is considered one of the most promising techniques to tackle tremendous high-dimensional data. This study attempts to extend deep learning theory into large-scale transportation network analysis. A deep Restricted Boltzmann Machine and Recurrent Neural Network architecture is utilized to model and predict traffic congestion evolution based on Global Positioning System (GPS) data from taxi. A numerical study in Ningbo, China is conducted to validate the effectiveness and efficiency of the proposed method. Results show that the prediction accuracy can achieve as high as 88% within less than 6 minutes when the model is implemented in a Graphic Processing Unit (GPU)-based parallel computing environment. The predicted congestion evolution patterns can be visualized temporally and spatially through a map-based platform to identify the vulnerable links for proactive congestion mitigation.
Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future ...motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs), for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs) and long short-term memory (LSTM) neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction.
This paper proposes a convolutional neural network (CNN)-based method that learns traffic as images and predicts large-scale, network-wide traffic speed with a high accuracy. Spatiotemporal traffic ...dynamics are converted to images describing the time and space relations of traffic flow via a two-dimensional time-space matrix. A CNN is applied to the image following two consecutive steps: abstract traffic feature extraction and network-wide traffic speed prediction. The effectiveness of the proposed method is evaluated by taking two real-world transportation networks, the second ring road and north-east transportation network in Beijing, as examples, and comparing the method with four prevailing algorithms, namely, ordinary least squares, k-nearest neighbors, artificial neural network, and random forest, and three deep learning architectures, namely, stacked autoencoder, recurrent neural network, and long-short-term memory network. The results show that the proposed method outperforms other algorithms by an average accuracy improvement of 42.91% within an acceptable execution time. The CNN can train the model in a reasonable time and, thus, is suitable for large-scale transportation networks.
•A Long Short-Term Memory Neural Network (LSTM) is developed for travel speed prediction.•The LSTM NN can capture the long-term temporal dependency for time series.•The LSTM NN can automatically ...determine the optimal time window.•A comparative study suggests that the LSTM NN receives the best performance.
Neural networks have been extensively applied to short-term traffic prediction in the past years. This study proposes a novel architecture of neural networks, Long Short-Term Neural Network (LSTM NN), to capture nonlinear traffic dynamic in an effective manner. The LSTM NN can overcome the issue of back-propagated error decay through memory blocks, and thus exhibits the superior capability for time series prediction with long temporal dependency. In addition, the LSTM NN can automatically determine the optimal time lags. To validate the effectiveness of LSTM NN, travel speed data from traffic microwave detectors in Beijing are used for model training and testing. A comparison with different topologies of dynamic neural networks as well as other prevailing parametric and nonparametric algorithms suggests that LSTM NN can achieve the best prediction performance in terms of both accuracy and stability.
•Monitoring mainland and island coastline variations of the Pearl River Estuary from 1978 to 2018.•Combined Otsu and Canny edge detections to extract coastline from Landsat dataset.•Reclassified ...coastline as biological, sandy, muddy, estuarine, rocky and artificial via visual interpretation.•Rezoned the study area into Area I-IV and analyzed coastline variations and differences among these areas.•Analyzed natural and anthropogenic factors of coastline variations of the study area.
The coastline of the Pearl River Estuary (PRE) experienced significant changes due to the increasing demand for land. Previous studies have focused on coastline changes of the mainland in China; thus changes of island coastlines have not been well investigated. In this study, the coastlines of the continent and islands of the Pearl River Estuary from 1978 to 2018 were extracted and analyzed by integrating Canny edge detection and Otsu methods based on Landsat data, and the coastline was classified into six types including biological, sandy, muddy, estuarine, rocky, and artificial via visual interpretation. The coastline length, coastline change rate and change area were calculated and analyzed for the 40 years period. The precision test showed that the coastline extraction methods presented in this paper meet the accuracy standard. The coastline length of the Pearl River Estuary was found to have increased rapidly from 789 km in 1978 to 979 km in 2018. During the past four decades, a large proportion of natural coastline converted into artificial one (66% in 2018), while 577 km2 land increased in this area. The coastline of the western PRE demonstrates more rapid change rates than that of the eastern part, and more land reclamation has occurred in the western area. Between 1988 and 1997, the coastline length and area of reclamation experienced the most extensive changes. Both natural factors, including sediment transport, sea-level rise, wind, tide, as well as anthropogenic factors, such as land reclamation and dam construction, have impacted coastline variation of the Pearl River Estuary in the past four decades. In general, the western part of the study area showed more remarkable coastline change and experienced the enormous influence of land reclamation, dam construction and sediment transport, compared with the eastern part.
Identifying and quantifying the influential factors on incident clearance time can benefit incident management for accident causal analysis and prediction, and consequently mitigate the impact of ...non-recurrent congestion. Traditional incident clearance time studies rely on either statistical models with rigorous assumptions or artificial intelligence (AI) approaches with poor interpretability. This paper proposes a novel method, gradient boosting decision trees (GBDTs), to predict the nonlinear and imbalanced incident clearance time based on different types of explanatory variables. The GBDT inherits both the advantages of statistical models and AI approaches, and can identify the complex and nonlinear relationship while computing the relative importance among variables. One-year crash data from Washington state, USA, incident tracking system are used to demonstrate the effectiveness of GBDT method. Based on the distribution of incident clearance time, two groups are categorized for prediction with a 15-min threshold. A comparative study confirms that the GBDT method is significantly superior to other algorithms for incidents with both short and long clearance times. In addition, incident response time is found to be the greatest contributor to short clearance time with more than 41% relative importance, while traffic volume generates the second greatest impact on incident clearance time with relative importance of 27.34% and 19.56%, respectively.
The coronavirus disease 2019 (COVID-19) epidemic has had a crucial influence on people's lives and socio-economic development throughout China and across the globe since December 2019. An ...understanding of the spatiotemporal patterns and influencing factors of the COVID-19 epidemic on multiple scales could benefit the control of the outbreak. Therefore, we used spatial autocorrelation and Spearman's rank correlation methods to investigate these two topics, respectively. The COVID-19 epidemic data reported publicly and relevant open data in Hubei province were analyzed. The results showed that (1) at both prefecture and county levels, the global spatial autocorrelation was extremely significant for the cumulative confirmed COVID-19 cases (CCC) in Hubei province from 30 January to 18 February 2020. Further, (2) at both levels, the significant hotspot and cluster/outlier area was observed solely in Wuhan city and most of its districts/sub-cities from 30 January to 18 February 2020. (3) At the prefecture level in Hubei province, the number of CCC had a positive and extremely significant correlation (
< 0.01) with the registered population (RGP), resident population (RSP), Baidu migration index (BMI), regional gross domestic production (GDP), and total retail sales of consumer goods (TRS), respectively, from 29 January to 18 February 2020 and had a negative and significant correlation (
< 0.05) with minimum elevation (MINE) from 2 February to 18 February 2020, but no association with the land area (LA), population density (PD), maximum elevation (MAXE), mean elevation (MNE), and range of elevation (RAE) from 23 January to 18 February 2020. (4) At the county level, the number of CCC in Hubei province had a positive and extremely significant correlation (
< 0.01) with PD, RGP, RSP, GDP, and TRS, respectively, from 27 January to 18 February 2020, and was negatively associated with MINE, MAXE, MNE, and RAE, respectively, from 26 January to 18 February 2020, and negatively associated with LA from 30 January to 18 February 2020. It suggested that (1) the COVID-19 epidemic at both levels in Hubei province had evident characteristics of significant global spatial autocorrelations and significant centralized high-risk outbreaks, and had an extremely significant association with social and economic factors. (2) The COVID-19 epidemics were significantly associated with the natural factors, such as LA, MAXE, MNE, and RAE, -only at the county level, not at the prefecture level, from 2 February to 18 February 2020. (3) The COVID-19 epidemics were significantly related to the socioeconomic factors, such as RGP, RSP, TRS, and GDP, at both levels from 26 January to 18 February 2020. It is desired that this study enrich our understanding of the spatiotemporal patterns and influencing factors of the COVID-19 epidemic and benefit classified prevention and control of the COVID-19 epidemic for policymakers.
Tumor immune microenvironment (TIME) include tumor cells, immune cells, cytokines, etc. The interactions between these components, which are divided into anti-tumor and pro-tumor, determine the trend ...of anti-tumor immunity. Although the immune system can eliminate tumor through the cancer-immune cycle, tumors appear to eventually evade from immune surveillance by shaping an immunosuppressive microenvironment. Immunotherapy reshapes the TIME and restores the tumor killing ability of anti-tumor immune cells. Herein, we review the function of immune cells within the TIME and discuss the contribution of current mainstream immunotherapeutic approaches to remolding the TIME. Changes in the immune microenvironment in different forms under the intervention of immunotherapy can shed light on better combination treatment strategies.
Electric storage units constitute a key element in the emerging smart grid system. In this paper, the interactions and energy trading decisions of a number of geographically distributed storage units ...are studied using a novel framework based on game theory. In particular, a noncooperative game is formulated between storage units, such as plug-in hybrid electric vehicles, or an array of batteries that are trading their stored energy. Here, each storage unit's owner can decide on the maximum amount of energy to sell in a local market so as to maximize a utility that reflects the tradeoff between the revenues from energy trading and the accompanying costs. Then in this energy exchange market between the storage units and the smart grid elements, the price at which energy is traded is determined via an auction mechanism. The game is shown to admit at least one Nash equilibrium and a novel algorithm that is guaranteed to reach such an equilibrium point is proposed. Simulation results show that the proposed approach yields significant performance improvements, in terms of the average utility per storage unit, reaching up to 130.2% compared to a conventional greedy approach.