Identification of travelers' transportation modes is a fundamental step for various problems that arise in the domain of transportation such as travel demand analysis, transport planning, and traffic ...management. In this paper, we aim to identify travelers' transportation modes purely based on their GPS trajectories. First, a segmentation process is developed to partition a user's trip into GPS segments with only one transportation mode. A majority of studies have proposed mode inference models based on hand-crafted features, which might be vulnerable to traffic and environmental conditions. Furthermore, the classification task in almost all models have been performed in a supervised fashion while a large amount of unlabeled GPS trajectories has remained unused. Accordingly, we propose a deep SE mi-Supervised C onvolutional A utoencoder ( SECA ) architecture that can not only automatically extract relevant features from GPS segments but also exploit useful information in unlabeled data. The SECA integrates a convolutional-deconvolutional autoencoder and a convolutional neural network into a unified framework to concurrently perform supervised and unsupervised learning. The two components are simultaneously trained using both labeled and unlabeled GPS segments, which have already been converted into an efficient representation for the convolutional operation. An optimum schedule for varying the balancing parameters between reconstruction and classification errors are also implemented. The performance of the proposed SECA model, trip segmentation, the method for converting a raw trajectory into a new representation, the hyperparameter schedule, and the model configuration are evaluated by comparing to several baselines and alternatives for various amounts of labeled and unlabeled data. Our experimental results demonstrate the superiority of the proposed model over the state-of-the-art semi-supervised and supervised methods with respect to metrics such as accuracy and F-measure.
This paper addresses a tactical joint inventory and transportation planning problem for multiple items with deterministic and time-varying demand, considering different transportation modes and item ...fragmentation. The latter corresponds to the splitting of the same item ordered quantity between several trucks or containers. On the one hand, fragmenting the items potentially reduces the number of containers used. On the other hand, loading the item lot fragments on several containers may negatively impact the handling and shipping operations. This new problem is proposed as a way to tackle such conflict. Several Mixed Integer Linear Programming models are proposed for the problem, which rely on two multi-item lot-sizing models with mode selection and two bin-packing models with item fragmentation. A relax-and-fix heuristic is also proposed. Using realistic instances, computational experiments are first conducted to identify the most efficient model in terms of computational time, to study the impact of key parameters on the computational complexity and to analyze the efficiency of the heuristic. Then, managerial insights are derived through additional computational experiments, in particular, to identify contexts requiring joint optimization of lot-sizing and bin-packing decisions, as well as the impact of item fragmentation constraints. Directions for future research are finally proposed.
The greenhouse gas emissions footprint and global warming potential are widely-used for environmental sustainability studies. However, environmental sustainability is far wider than carbon emissions ...and climate change. This review aims to highlight the importance of considering air pollutants in optimisation studies and evaluate the limitation of the current assessments for air emissions, particularly in relation to transportation. The source of air pollutants is firstly overviewed with special attention on non-stationary sources, freight and sea transportation. The type of measurement to obtain the emission data and the available optimisation models on transport mode choice selection were then summarised. The strengths and Weaknesses' have been indicated. The identified gap includes greenhouse gas and air pollutants not being evaluated simultaneously and the interaction between the different pollutants are not being adequately considered. A better assessment framework and impact categories classification are consequently required. The summarised assessment model of transportation mode choice shows that the current viewpoint on low emissions, green or environmental sustainability options refers to carbon dioxide as a part of greenhouse gas. Attention towards a better emission assessment and management has been supported in this study through critical discussion. The next step of this work is to develop a methodology to measure greenhouse gas and air pollutants simultaneously by considering the synergistic effect and the discussed limitation. It is important for minimising the potential of footprint shifting and poor decision-making.
•A CNN architecture is proposed to infer transportation modes from GPS trajectories.•An adaptable and efficient layout for the input layer of the CNN is designed.•Key factors in the CNN: remove ...anomalies, data augmentation, use the bagging concept.•The proposed CNN achieves the accuracy of 84.8%, higher than other studies.
Identifying the distribution of users’ transportation modes is an essential part of travel demand analysis and transportation planning. With the advent of ubiquitous GPS-enabled devices (e.g., a smartphone), a cost-effective approach for inferring commuters’ mobility mode(s) is to leverage their GPS trajectories. A majority of studies have proposed mode inference models based on hand-crafted features and traditional machine learning algorithms. However, manual features engender some major drawbacks including vulnerability to traffic and environmental conditions as well as possessing human’s bias in creating efficient features. One way to overcome these issues is by utilizing Convolutional Neural Network (CNN) schemes that are capable of automatically driving high-level features from the raw input. Accordingly, in this paper, we take advantage of CNN architectures so as to predict travel modes based on only raw GPS trajectories, where the modes are labeled as walk, bike, bus, driving, and train. Our key contribution is designing the layout of the CNN’s input layer in such a way that not only is adaptable with the CNN schemes but represents fundamental motion characteristics of a moving object including speed, acceleration, jerk, and bearing rate. Furthermore, we ameliorate the quality of GPS logs through several data preprocessing steps. Using the clean input layer, a variety of CNN configurations are evaluated to achieve the best CNN architecture. The highest accuracy of 84.8% has been achieved through the ensemble of the best CNN configuration. In this research, we contrast our methodology with traditional machine learning algorithms as well as the seminal and most related studies to demonstrate the superiority of our framework.
•Explore the value of Internet based elastic logistics platforms (ELPs).•Study how ELPs enhance fashion quick response systems.•Derive via stochastic dynamic programming the optimal transportation ...option and ordering policy.•Examine the situations in which ELPs are especially beneficial to apply.
Quick response (QR) systems, which aim to respond promptly to market changes by postponing inventory decisions with a reduced lead time, are critical to support fast fashion operations. Under QR systems, in the digital data analytics era, the fashion brand can acquire market information using digital technologies to improve demand forecasting. This enhances the respective ordering and transportation mode selection decisions. It is known that to implement QR systems requires the availability of the right logistics option (e.g., transportation mode) and the needed logistics capacity in the future. An Internet based elastic logistics platform (ELP) hence emerges to help. In this paper, in the main model, we analytically derive via stochastic dynamic programming the optimal transportation option selection and inventory ordering policy with the ELP. We further examine the value of ELP and identify the situations in which it is especially beneficial to adopt it. Robustness checking is conducted which proves that the theoretical findings derived in the main model are solid. Specific managerial action plans and managerial implications are developed.
Access to subway stations is important for daily commuting, but scant attention has been given to green space exposure at subway station areas in people’s residential neighbourhoods and workplace ...areas. This paper focuses on the association between street-level green space exposure around subway stations at residential and work locations and people’s choice of subway as their primary commuting mode and travel satisfaction, using street view data and survey data in Beijing, China. Street view data and a machine learning approach were used to measure both street view green space quantity (SVG-quantity) and street view green space quality (SVG-quality). The results suggested that SVG-quantity and SVG-quality generate differential effects on subway use and travel satisfaction under residential and workplace contexts. Findings of this study highlight the complementary effects of green space and travel infrastructure provision in shaping travel behaviour and wellbeing in residential neighbourhood and workplace contexts.
Hydrogen energy has the advantages of being clean and carbon-free, convenient storage, and easy conversion to electrical energy. The use of renewable energy electrolysis for hydrogen production is an ...important solution to promote clean energy utilization and decarburization in the power and transportation industries. However, hydrogen has a low hydrogen storage density and is prone to explosion. The high cost and risk of hydrogen transportation pose challenges to the promotion and utilization of hydrogen. To address these difficulties, this paper proposes a day-ahead dispatch of electricity-hydrogen systems (EHS) under the solid-state transportation mode of hydrogen energy. Firstly, based on the Van't Hoff equation, the relationship between gas pressure and the reaction temperature is established during the gas-solid conversion process, and a hydrogen energy solid-state transport model based on a magnesium-based hydrogen transport vehicle (MHTV) is proposed. Secondly, a renewable energy uncertainty set is established based on the information-gap decision theory (IGDT) envelope constraint for addressing the uncertainty of renewable energy and a day-ahead bi-level dispatch model of the EHS is constructed based on the solid-state transportation mode of hydrogen energy. Finally, the IGDT approach considering fuzzy variables (FV-IGDT) is presented by using a fuzzy variable membership function, and the proposed bi-level model is transformed into a single-level model for a solution based on this approach. The effectiveness of the proposed model and solution method is verified through simulation on an EHS consisting of a modified IEEE-118 power system, 2 hydrogen production stations (HPS), and 10 hydrogen refueling stations (HRS).
•Hydrogen pressure & temperature during gas-solid conversion established by Van't Hoff equation.•Hydrogen energy solid-state transport model based MHTV is proposed using magnesium.•MHTV hydrogen transportation time, energy balance, time-coupling constraints are considered.•FV-IGDT model is proposed to address issue of traditional IGDT models unable to solve optimal value.•Optimize fluctuation range of uncertain variables in RSOS and RARS models and optimal values.
This paper adopts different supervised learning methods from the field of machine learning to develop multiclass classifiers that identify the transportation mode, including driving a car, riding a ...bicycle, riding a bus, walking, and running. Methods that were considered include K-nearest neighbor, support vector machines (SVMs), and tree-based models that comprise a single decision tree, bagging, and random forest (RF) methods. For training and validating purposes, data were obtained from smartphone sensors, including accelerometer, gyroscope, and rotation vector sensors. K-fold cross-validation as well as out-of-bag error was used for model selection and validation purposes. Several features were created from which a subset was identified through the minimum redundancy maximum relevance method. Data obtained from the smartphone sensors were found to provide important information to distinguish between transportation modes. The performance of different methods was evaluated and compared. The RF and SVM methods were found to produce the best performance. Furthermore, an effort was made to develop a new additional feature that entails creating a combination of other features by adopting a simulated annealing algorithm and a random forest method.
•Raw water pipelines can remove a certain extent of long-chain PFASs.•Long-distance, open-channel water diversion can increase PFAS contamination risk.•O3-BAC and UV treatment processes can provide a ...certain removal effect on PFASs.•The PFAS distribution is affected by the loose deposit in branched DWDSs.•The PFAS concentration forms a relatively uniform distribution in the looped DWDSs.
Perfluoroalkyl substances (PFASs) can occur in water sources, pass through drinking water treatment plants (DWTPs), drinking water distribution systems (DWDSs), to the consumer taps. This investigation was carried out to present the transportation behaviors of 17 PFASs, involving seven DWTPs with different water sources, raw water transportation modes, treatment processes, and DWDS structures in eastern and northern China. The results showed that the long-distance raw water transportation pipelines removed a certain extent of PFASs from raw water, probably due to the accumulation of loose deposits. The long-distance, open-channel South-to-North water diversion increased PFAS contamination risk. In the DWTPs, granular activated carbon (GAC) adsorption and ultraviolet radiation removed less than 25% of PFASs, but ozonation-biological activated carbon (O3-BAC) was superior to GAC alone in removing PFASs. Loose deposits couldsignificantly influence PFAS accumulation and release within branch-structured DWDSs. In loop-structured DWDSs, finished water with different PFAS characteristics could mix along the pipeline, with the corresponding DWTP as the center, ultimately forming a relatively uniform distribution in the entire DWDS.
Display omitted