交通信息与安全

2023, 41(2): .

Abstract(428) HTML (303) PDF(41)

Abstract:

A Review on the Safety Studies at Entrances and Exits of Expressway Interchanges

ZHANG Chi, LIU Kai, WANG Shifa, XIE Zilong, WANG Xue

2023, 41(2): 1-17. doi: 10.3963/j.jssn.1674-4861.2023.02.001

Abstract(1195) HTML (382) PDF(110)

Abstract:
Interchanges of expressways are critical nodes for traffic operation, and their entrances/exits are key components and show most complex traffic flow behaviors. Researchers have conducted numerous safety studies on interchanges and obtained a variety of conclusions, but their studies are not systematic due to the complexity of interchanges and the entrances/exits. To this end, this paper looks into the literature on the safety of interchanges and entrances/exits of expressways, identifies corresponding research topics of interest, and reviews corresponding findings in three areas: geometric design index, safety evaluation method and operation speed. Study results indicate that: ①The number of factors considered during the design of speed-change lanes in the existing standards in China is not enough to match the required distance for speed change from the field. The criterion of sight distance at the merging end of an interchange is not optimized based on vehicle performance and drivers' perception-reaction time to different traffic scenarios. ②Studies in the literature on the merging nose are mainly carried out from a two-dimensional perspective, and the geometric design factors are not focused on the effect of the combination of flat and longitudinal factors. ③The accident statistics method, analytic hierarchy process, and fuzzy comprehensive evaluation method are most commonly used in evaluating the safety of expressways. The accident statistics method can objectively be used to evaluate the safety, but it is dependent on the accuracy of accident statistics; the analytic hierarchy process can be used to develop safety evaluation indexes, but the weights associated to the indexes are often subjectively determined, which often makes the results useless; the fuzzy comprehensive evaluation method relies on the selection of the affiliation function to ensure a good accuracy of the evaluation of the level of safety, which places high requirements on the knowledge of the user. ④The evaluation criteria of coordination and continuity of road alignment in the existing standards are based on the traditional method Δv₈₅, whose basis is contrary to the reality leading to overestimating the safety performance of the road. Therefore, speed prediction models for different vehicles in different entrances/exits should be studied in depth. From these results, the suggestions for the future works are as follows: ①Studies on speed-change lanes can be based on driving behavior, with a focus on the analysis of acceleration patterns over speed-change lanes. ②Studies on sight distance at entrances/exits should take into account complex traffic circumstances, drivers' reaction time under visual stimuli and characteristics of vehicles. Further optimization can be conducted based on the unique characteristics of entrances/exits, and tested with relatively mature safety evaluation methods, such as accident statistics methods and prediction models. ③In terms of the operating speed at entrances/exits, floating vehicle data (FCD) can be used together with the data from gantry cranes and radars to study the process that vehicles change their speed.

A Method for Timely Detecting Foreign Objects between Metro Platform Screen Doors and Train Doors Based on an Improved YOLOv5s Model

DAI Yuan, LIU Weiming, WANG Heng, XIE Wei, LONG Kejun

2023, 41(2): 18-27. doi: 10.3963/j.jssn.1674-4861.2023.02.002

Abstract(965) HTML (429) PDF(91)

Abstract:
Accurately and efficiently detecting foreign objects between platform screen doors (PSDs) and train doors at metro stations is of great significance for safety purpose. In response to the inefficiency and inaccuracy of current detection methods, a method based on the you-only-look-once (YOLOv5s) model is proposed. As the original YOLOv5s model relies on internal features of candidate regions but not global contextual information, a global context module is introduced to address the limitation. This module integrates non-local modules and squeeze-excitation modules. The non-local modules use self-attention mechanism to model relationships between pixels and capture long-term dependencies. The squeeze-excitation modules is developed to reduce the computational cost of the model. The global context module enables the model to capture global contextual information and combines it with local information for improved detection of foreign objects without significantly increasing computational complexity. Additionally, the inefficient Focus module of the original YOLOv5s is replaced with a Stem module that is fully developed from standard convolutional units, contributing to a reduced computation cost and enhanced detection speed. Experiments are conducted based on a dataset of 5 854 foreign object images collected from metro stations, with the model being tested using desktop-level NVIDIA TITAN Xp graphics cards. The results indicate that ①the improved YOLO model performs remarkably better than other baseline models, exhibiting an impressive detection speed of 385 frames per second, a 100% improvement over the original YOLOv5s model and a substantial 466% improvement over the fastest speed of YOLOv3-SPP model. ② The improved YOLO model achieves an average detection accuracy of 88.5%, a 0.5% improvement over the original YOLOv5s and a 0.6% improvement over the highest average detection accuracy of YOLOv3-SPP. ③ The improved YOLO model takes up only 14.4 MB of computer storage space, which is 0.7% less than the original YOLOv5s, and 85% less than the single shot multibox detector (SSD) that takes the least storage space.

Effects of Spacing of Highway Roadside Millimeter-wave Radar Detectors on the Accuracy of a Crash Risk Evaluation Model

YANG Dongfeng, DAI Jie, ZHANG Yueyan, HAN Lei, YU Rongjie

2023, 41(2): 28-35. doi: 10.3963/j.jssn.1674-4861.2023.02.003

Abstract(931) HTML (430) PDF(53)

Abstract:
Freeways equipped with new sensing equipment such as millimeter-wave radar detectors can accurately monitor traffic operation and well support active traffic management measures. However, due to the high deployment expenditure, there is a need to consider the cost constraints and the effectiveness of traffic state detection. To investigate the impacts of millimeter-wave radar deployment spacing on crash risk evaluation performance, this study is conducted based on the empirical data of the Hangshaoyong highway in Zhejiang Province. A crash risk evaluation model based on deep forest (DF) is developed. Specifically, sliding spatio-temporal windows are employed to extract the features of traffic operation while the correlation relationships between the features and crash risk are established through the integrations of multi-layer cascaded random forests. Considering the sensing range of the millimeter-wave radar detectors, multiple traffic operation datasets are developed by assuming different deployment spacings. Sensitivity analyses of radar deployment spacing on the evaluation accuracy of crash risk are conducted. Analyses results show that: The area under curve (AUC) of DF model is 0.849 with 80.9% recall on crash samples, which is higher than traditional convolutional neural network model (AUC is 0.741, recall is 75.2%) and random forest model (AUC is 0.715, recall is 70.8%). An inverse relationship between radar deployment spacing and evaluation accuracy of crash risk is captured, and the marginal effects of the improvement to the model accuracy decreases under dense deployment conditions. If the radar deployment spacing is reduced from 1 500 m to 750 m, the AUC of crash risk evaluation model shows a substantial increase (from 0.794 to 0.853), but there is no obvious change in AUC values when the radar deployment spacing is reduced from 750 m to 250 m. In conclusion, the radar deployment spacing of 750 m can balance the deployment cost and the evaluation performance of crash risk, which could be used to support the decisions related to the installment of traffic sensing equipment.

A Prediction Method of Daily Traffic Accident Frequency at Black Spots Based on Bi-Directional Long Short-term Memory Networks

REN Yi, YANG Renfa, ZHOU Jibiao, HU Zhenghua, ZHANG Minjie

2023, 41(2): 36-49. doi: 10.3963/j.jssn.1674-4861.2023.02.004

Abstract(1000) HTML (391) PDF(60)

Abstract:
It is of significance to provide timely warning of accidents to traffic management departments (TMDs) and the public. Therefore, a method that predicts traffic accident frequency at black spots is proposed, using a bidirectional long short-term memory neural network (BiLSTM). An improvement in k -value selection within the traditional K-means clustering algorithm increases the efficiency of identifying black spots and the number of accidents at each of black spots is collected and used to develop an accident time series dataset. The wavelet decomposition is used to denoise the time series; a multi-layer grid search method is employed to calibrate the model parameters such based on a BiLSTM network. Traffic flow, holidays, weather, and accident environment are set as external parameters of the model and the internal parameters are obtained from the accident time series via a sliding window method. Moreover, the frequencies of accidents at black spots in the next day are predicted based on the internal and external parameters, and then an early warning model for traffic accident at black spots is proposed based on the prediction; Lastly, the accident data collected by the TMD in the City of Ningbo, Province of Zhejiang from April 2020 to September 2021 are used as the test set, the frequencies of accidents at black spots in the next day are predicted based on the accident data from the previous 7 days, and the proposed BiLSTM model is compared with other prediction models such as gated recurrent unit (GRU), long short-term memory (LSTM), back propagation (BP) neural network, autoregressive integrated moving average (ARIMA), and support vector regression (SVR). Study results show that the average accuracy of the BiLSTM, GRU, LSTM, BP, ARIMA, and SVR models for predicting daily accident frequencies at each black spots is 93.1%, 88.8%, 88.0%, 85.2%, 84.4%, and 84.2%, respectively, and the root mean square error of the above models is 0.092, 0.146, 0.142, 0.147, 0.177, and 0.176, respectively. Study results indicate that the proposed method has a higher prediction accuracy and better robustness than the traditional methods. as the number of hidden layers and nodes, etc., and a prediction model for the accident frequency is developed

Influence of Horizontal and Vertical Alignments of Undersea Tunnel on Driver's Visual Characteristics and Vehicle Speed

PAN Fuquan, YANG Jingzhou, ZHANG Lixia, LI Weicong, YANG Xiaoxia, BING Qichun

2023, 41(2): 50-58. doi: 10.3963/j.jssn.1674-4861.2023.02.005

Abstract(867) HTML (397) PDF(29)

Abstract:
The combination of horizontal and vertical alignments of the undersea tunnel is complex and diverse, which can easily lead to adverse reactions such as distraction and fatigue for drivers. To this end, the data of 30 drivers are collected to quantify and analyze the effect of the cross-harbor tunnel flat and longitudinal alignment on drivers' visual characteristics and vehicle speed. The Facelab eye tracker, GPS X10 vehicle-mounted inclinometer, tachograph, and other equipment are used to collect the drivers' blink frequency, percentage of eye closure over the pupil per unit time (PERCLOS), tunnel slope, vehicle speed, and other data. The correlation and significance between the slope and curvature of the undersea tunnel and the drivers' blink frequency, PERCLOS, and vehicle speed are analyzed using partial correlation analysis. Then, Ploy 2D nonlinear surface fitting is used to establish mathematical models of blink frequency, PERCLOS, and vehicle speed with slope curvature. The relationship between blink frequency, PERCLOS, vehicle speed, and the horizontal and vertical alignments of the undersea tunnel is quantitatively analyzed, thus reflecting the influence of different combinations of horizontal and vertical alignments of the undersea tunnel on the driver's mental state and driving condition. Study results indicate that: Under the combination of a slope of 1.3% and the radius of the circular curve of 4 348 m, the drivers' blink frequency is highest, and their mental state is most relaxed. The driver's tension will be increased by appropriately increasing the positive slope and decreasing the curvature radius. Under the combination of a slope of 3.05% and the radius of the circular curve of 3 521 m, the drivers' PERCLOS is the largest and the fatigue degree is the highest. By appropriately increasing the negative slope and reducing the curvature radius the driver's tension will be relieved. Under the combination of a slope of 1.78% and the radius of the circular curve of 2 433 m, the vehicle speed is found to be the highest. The vehicle's speed will be reduced by appropriately increasing the positive slope and curvature radius. The constructed visual characteristics and speed model can reflect the changes in the driver's mental state and driving condition with the changes in horizontal and vertical alignments, which shall provide theoretical support for the horizontal and vertical alignments' safe design and operation management of the undersea tunnel.

Night vehicle detection method based on improving Mask RCNN

LIU Jie, JIN Jide, ZHENG Qingxiang

2023, 41(2): 59-66. doi: 10.3963/j.jssn.1674-4861.2023.02.006

Abstract(833) HTML (435) PDF(37)

Abstract:
The traditional nighttime vehicle detection method is generally based on the extraction and identification of headlights, which is prone to misjudgment, detection accuracy, and real-time lack of deficiency. To address the above issue, a night vehicle detection algorithm based on improved Mask RCNN night vehicle detection (Mask RCNN) is studied. The normal convolution in the residual network (ResNet) structure is modified to a grouped convolution of 16 groups, and channel number superposition is achieved by 16 groups of 1×1 convolution Network parameters are reduced to 1/16 of a normal convolution. The detection speed is improved and achieves the same effect as normal convolution. The channel attention mechanism module (squeeze-and-excitation, SE) is embedded in the ResNet structure, two fully connected layers are used to build the bottleneck structure, normalized weights are weighted to each channel feature, The representational power of the network is enhanced; Bottom-up structures are added behind feature pyramid networks (FPNs), the strong localization information of the underlying features is passed to the high-level semantic features; Adaptive pooling layer is added. The region proposal network (RPN) generates candidate regions, which are subsequently assigned to feature maps at different scales by region. The jump connection structure is added between the bottom feature and the top feature at each stage, The model parameters are reduced, while retaining the global representational power of the model. The open-source dataset Microsoft common objects in context (MS COCO), and Berkeley Deep Drive 100K (BDD100K) contain some nighttime driving images, these images are image enhanced. A test set is constructed to evaluate the performance of the detection, it contains 2 000 images. The Mask RCNN-NVD algorithm is tested on the test set, the mean Average Precision of Mask RCNN-NVD is 92.62, and the Frames Per Second (FPS) of Mask RCNN-NVD is 30 frames. Compared with the original Mask RCNN algorithm, the mapped value is improved by 1.68 and the FPS value is improved by 4 frames. The proposed method is validated, and nighttime vehicle detection is improved in both accuracy and real time.

Modeling Car Following Behavior of Autonomous Driving Vehicles Based on Deep Reinforcement Learning

CHEN Yue, JIAO Pengpeng, BAI Ruyu, LI Rujian

2023, 41(2): 67-75. doi: 10.3963/j.jssn.1674-4861.2023.02.007

Abstract(1004) HTML (337) PDF(109)

Abstract:
In order to enhance the performance of car following behavior of autonomous vehicles and mitigate the negative effects of traffic oscillations, a deep reinforcement learning-based car following model for automated driving is investigated. The existing reward function is improved by incorporating energy consumption, and the related terms for representing energy consumption are established based on the VT-Micro model. In addition, the method of using the time gap between vehicles to establish the reward function related to driving efficiency is improved by adding virtual speed to the time gap, in order to avoid computation overflow and unrealistic short following distance in the traffic oscillation scenario. To overcome the limitations of training on closed-loop simulated roads and simulated vehicle trajectories, human driver behavior extracted from the NGSIM trajectory data during traffic oscillation are used to develop the training environment. By applying the twin delayed deep deterministic policy gradient algorithm (TD3), a multi-objective car following model is then developed. A system for evaluating model performance is established to compare the performance of the TD3 model with traditional models in car following and traffic oscillations scenarios. Study results of car following scenarios show that the TD3 model and the traditional adaptive cruise control (ACC) model perform similarly in terms of comfort and driving efficiency, but both outperform the human drivers. In terms of safety, the TD3 model reduces safety hazards by 53.65% compared to the traditional ACC model, and 36.24% compared to the human drivers. Regarding energy consumption, the TD3 model reduces the energy consumption of the conventional ACC model and human drivers by 6.73% and 15.65%, respectively. Study results show that the TD3 model can reduce the negative impacts of traffic oscillations. In the scenario with a 100% TD3 model penetration rate, driving discomfort decreases by 55.95%, driving efficiency increases by 8.82%, crash risks reduce by 73.21%, and fuel consumption drops by 5.97%, compared to a 100% human-driven environment.

A Method for Monitoring Traffic State in the Airspace Based on an Improved Support Vector Machine

ZHU Chengyuan, ZHANG Che, GUAN Jianhua

2023, 41(2): 76-85. doi: 10.3963/j.jssn.1674-4861.2023.02.008

Abstract(778) HTML (345) PDF(29)

Abstract:
This paper quantitatively analyzes the methods for monitoring airspace traffic state from the perspective of the workload of air traffic controllers, due to the difficult measurement of such a factor in current studies.In response to the need of monitoring airspace traffic state more efficiently, an airspace simulation model is developedbased onTotal Airspace and Airport Modeller (TAAM) software and a method for identifying traffic state in the rasterized airspace scenario is proposed based on an improved support vector machine (SVM). Based on real-world operation experience of controllers, the shape and size of grids are determined by comparing different rasterization schemes. Target airspace is rasterized by taking the hexagon with a side length of 25 km as the smallest unit. Considering a variety of controller's workloads and the distribution of navigation facilities, a set of indicators for describing traffic states of the airspace is developed.Ak-means clustering algorithm is used to generate prior classified data by aggregating simulated sample data. A traffic state model for the airspace, developed based on the sparrow-search algorithm (SSA) and SVM, iscalled SSA-SVM.The solution set is divided according to the fitness. Moreover, key parameters of the model, including kernel parameters σ and penalty coefficients C, are optimized to determine a combination of parameters, which can increase the generalization capability of the model and avoid overfitting. Traffic states in the rasterized airspace are divided into four levels. Simulations are conductedfor the control airspace ofthe City of Xi'an. Study results show that the proposed SSA-SVM model can mitigate the overfitting problem, but not by the proposed genetic algorithm and support vector machine (GA-SVM) model.The average accuracy of classification is improved by 2.50%, and the accuracy of classification is improved by 1.73%. Among the tested 176 grids, the number of congested, crowded, and steady grids are 26, 18, and 51, respectively. Compared with the partition method for the complex airspace based on controller experience, the convergencerate of the proposed model is as high as 95%, which verifies the effectiveness of the proposed method for identifying airspace traffic state and reducing the workload of air traffic controllers.

A Method of Generating Effective Paths of Urban Rail Transit Based on Passenger Fare Collection Data

YIN Shisong, LU Bincheng, YE Mao, YANG Zhiqiang

2023, 41(2): 86-94. doi: 10.3963/j.jssn.1674-4861.2023.02.009

Abstract(828) HTML (350) PDF(26)

Abstract:
It is a fundamental task to generate effective paths for predicting cross-section passenger flows, calculat-ing network capacity, and analyzing passenger demand of urban rail transit system. To solve the problems of tradi-tional effective path generation in which the validity of each path cannot be evaluated, and the linear constraint can-not be assigned, as well as aiming at reducing the influences of randomness of questionnaire on the final path set generation, a method of generating effective path sets is developed based on traditional survey data. The proposed method analyzes route choice behavior of passengers and makes corresponding hypothesis. Then, models for evalu-ating effective routes under different clusters are established by introducing passenger trip duration. Additionally, the stations and the lines between them are abstracted as nodes and edges of the network of rail transit. And then, considering route types, subjective factors, and the density of passenger flows, the passenger's travel time data are processed using the adaptive DBSCAN algorithm and are divided into clusters according to the density of passenger flows at different time intervals. Furthermore, taking the clustering results as the input, a Logit model is developed to replace the linear constraints in path generation. Weights of effective paths represented by clusters are calculated separately, and the effective path set is obtained based on its continuity characteristics at travel time intervals. Exper-iments are conducted based on multiple origin-destination trip data from the metro network in the City of Guang-zhou. Study results show that the effective paths have an average adjusted Rand index of 0.652. Compared with the traditional algorithms, the adjusted Rand index has improved by 0.379. It indicates that the proposed method produc-es a smoother set boundary in the plane of route length and transfer time, which is more adaptable to complex and changeable networks.

A Method for Predicting Speed of Vehicles on Expressways Based on a ST-GCAN Model

YANG Peihong, XU Yanjun

2023, 41(2): 95-102. doi: 10.3963/j.jssn.1674-4861.2023.02.010

Abstract(769) HTML (293) PDF(22)

Abstract:
The speed of vehicles on expressways is a significant indicator for describing the effectiveness and safety of road transportation system. Accurate prediction of vehicle speed on expressways can contribute to reduction of traffic accidents and improvement of the level of services. In this sense, a prediction method for vehicle speed, called ST-GCAN, is developed, which integrates graph convolutional neural network (GCN), long short-term memo-ry network (LSTM) and attention mechanism into one model. Graph convolutional network is used to extract the spatial correlations of complex networks of expressways, long-short term memory network is used to extract the temporal correlations of historical data of vehicle speed, and attention mechanism is used to aggregate and analyze the correlation between historical data and predicted vehicle speed. In addition, the model employs dense connec-tions and layer normalization to ensure the integrity of information in the prediction model and to solve the problem of covariate shift during training. The model is tested with a dataset of vehicle speed on expressways of the City of Xining, Province of Qinghai, which contains a total of 94 777 hourly observations on 49 road sections at 8 toll sta-tions from May 1 to August 31, 2020. The ST-GCAN model predicts the vehicle speed in the next hour withthe mean absolute error (MAE) of 12.762%, root mean square error (RMSE) of 21.535%, and R² of 0.651.Compared to the HA model and the ARIMA model, the MAE of the ST-GCAN model is reduced by 11.1% and 19.7%, respec-tively. Compared to other deep learning models, it is reduced by approximately 8.0% to 10%. In conclusion, the ST-GCAN model can accurately estimate vehicle speed on expressways and shall be able to meet the requirements of intelligent traffic control systems.

A Method for Identifying Operation Zones of Free-floating Shared Bikes Based on Leiden Algorithm: A Case Study of the City of Nanjing

CHENG Cheng, CHEN Wendong, MA Hongsheng, LIU Xize, CHEN Xuewu

2023, 41(2): 103-111. doi: 10.3963/j.jssn.1674-4861.2023.02.011

Abstract(823) HTML (293) PDF(46)

Abstract:
In the operation and management of the free-floating shared bike (FFSB) industry, the operation zones are mainly determined based on administrative boundaries of districts without fully considering the spatial distribu-tions of travel demand of FFSB, resulting in a large number of inter-zone transfer tasks which seriously deteriorates the efficiency of its operation. To this end, a new method for identifying operation zones of FFSB based on a Leiden community detection algorithm is developed using the bike order data from the City of Nanjing. A three-layer data structure of"travel OD (origin-destination)-traffic zone-spatial interaction network"is developed. The Leiden com-munity detection algorithm is used to identify the FFSB communities, which are taken as the operation sub-zones of FFBS to divide the operation zones. By comparing the communities of FFBS in different years, the temporal charac-teristics of the spatial distribution of FFBS travel are revealed. In addition, two indicators, network modularity and computational efficiency, are adopted to compare the performance of various community detection algorithms and to further verify the effectiveness and superiority of the Leiden algorithm in this research problem. The results show that: ①regarding the FFBS travel in 2019, the proposed algorithm identifies 23 activity communities, and the pro-portion of FFBS travel within the communities reaches 82.9%, which is higher than the traditional partition method by 11%. This indicates that the proposed algorithm can make more FFBS travel be classified within communities, in-crease the self-cycle rate of shared bikes within a zone, and improve the operational efficiency. ② Comparing to the case in 2019, the scale of communities decreased and the number of communities increases in 2022, implying a re-duction in the travel distance of FFBS users and a decrease in the proportion of inter-zone travel. ③ In terms of the results from the proposed algorithm, the network modularity reaches 0.55, which is significantly improved, compar-ing with the results of traditional CNM algorithm (0.2), Walktrap algorithm (0.31) and Louvain algorithm (0.42). The computation time of the proposed algorithm is 1.1 s, while for the other three algorithms, this value is 6.4, 1.6, and 1.4 s, respectively, which demonstrates the proposed algorithm has a significant improvement in computing effi-ciency. The above results show that the Leiden algorithm is superior to others in terms of partition quality and com-putational efficiency. The proposed method reveals that the spatial characteristics of FFBS travel and can obtain a better zonal management scheme for FFBS, which provides the theoretical guidance for the reasonable determina-tion of partition operation schemes of FFBS.

An Analysis of Impact Factors to the Propensity of Car Sharing in the "Post-Epidemic Era"Based on an Extended UTAUT Model

WU Wenjing, YANG Xu, JIA Hongfei

2023, 41(2): 112-120. doi: 10.3963/j.jssn.1674-4861.2023.02.012

Abstract(672) HTML (283) PDF(26)

Abstract:
In the post-epidemic period, car sharing offers a new choice of transportation mode because of its "low-carbon"attributes and independent space for travelers. A survey of travelers is created in the form of a web-based questionnaire. A total of 109 valid questionnaires were collected and the results are analyzed to investi-gate factors impacting people's propensity to use car sharing in the post-epidemic period. An expanded UTAUT model is developed by considering the perceived risk of epidemic and financial risk. To investigate how each latent variable influences the intention to accept car sharing, eleven hypotheses are proposed, and structural equation mod-eling is applied. Analysis of the hypothesis test findings, model degree of fit, and mediating effects are done. Mediat-ing effects are looked at for routes that are insignificant in the hypothesis. A multiple factor and multiple reasoning model based on structural equations is developed to examine the effect process of socio-economic variables. Correla-tions between observed and latent variables are examined, in addition to correlations between latent variables and la-tent variables. Findings from this study show that that all of the models have a good fit. Performance expectation had the greatest, positive influence on acceptance intention among the latent variables, followed by facilitation con-ditions and social influence. Financial risk and effort expectation have the greatest negative influence on acceptance intention. Social influence, performance expectation, and facilitation conditions partially mediated the effect be-tween perceived risk of epidemic and behavioral intention. This effect had a total indirect effect of 0.240 and a sig-nificant indirect effect of 74.8% of the total effect. However, the direct effect of perceived epidemic risk is not signif-icant. Age, actual driving age, and the possession of a driver's license all significantly influence the attitudes regard-ing the use of car sharing during the pandemic. In contrast, frequency of usage directly affectes intention to use car sharing during the epidemic. Strategies and instructions for promoting car sharing in the post-epidemic era are then offered, which include ways to improve the travel experience, strengthen safety management, encourage consump-tion, and boost brand value.

A Method of Predicting Short-term Traffic Flows Based on a DGC-GRU Model

CUI Wenyue, GU Yuanli, ZHAO Shengli, RUI Xiaoping

2023, 41(2): 121-128. doi: 10.3963/j.jssn.1674-4861.2023.02.013

Abstract(731) HTML (273) PDF(33)

Abstract:
In order to study the spatiotemporal characteristics of traffic flows on urban expressways which have not been fully explored in previous studies, a shortterm prediction method for traffic flow on urban expressways is proposed, in order to improve the prediction accuracy and efficiency based on a combined model of directed graph convolutional neural networks and gated recurrent units (DGC-GRU). The proposed method uses a spatial correlation matrix, which is also combined with a graph convolutional neural network. A directed graph convolutional neural network (DG-CNN) is developed to characterize the directionality and variability of traffic flows. Traffic flow parameters are input into the DG-CNN to obtain the directed graph convolution operator, and the directed graph convolution operator is introduced into the gated loop unit. The DG-CNN and the gated loop unit are used to capture spatial and temporal features of traffic flow, respectively, and are combined to predict traffic flow on expressways. Traffic flow data collected from a Ring Expressway of the City of Seattle is used for experiment analysis, in order to compare the performance of the proposed prediction models. Study results show that the convergence speed of the proposed DGC-GRU model is faster than other baseline models, and the mean absolute error (MAE) and mean absolute percentage error (MAPE) of the DGC-GRU model are smaller than those of the baseline models, given the same dataset and parameter settings. Compared with traditional GRU, GCN, and DGC-LSTM models, the DGC-GRU model reduces the MAE by 33.01%, 5.76%, 1.32%, and MAPE by 27.75%, 11.15%, 7.76%, respectively, which indicate that the DGC-GRU model can effectively study the spatiotemporal characteristics of traffic flows of the urban expressway network and has a better performance on prediction accuracy and efficiency than the compared models.

A Method for Developing Operation Plans of Community Railways under the Condition of Multi-network Integration

LI Yina, MENG Xuelei, QIN Yongsheng, HAN Zheng, WANG Yue

2023, 41(2): 129-138. doi: 10.3963/j.jssn.1674-4861.2023.02.014

Abstract(566) HTML (216) PDF(13)

Abstract:
A method for developing operation plans of community railways under the condition of multi-network integration is studied, in order to match arrival-departure time sequence and capacity between community railways and other types of railways. The flows of cross-mode transfer are considered and divided according to its time distribution and the value of time of passengers. A multi-objective nonlinear integer programming model, whose objective is to minimize train operation cost, generalized total travel cost of passengers, and the short-term impact of cross-mode transfer on the studied railway line, is developed with the constraints of load rate, capacity, time interval, arrival-departure time of trains and coordination of transfers of passengers. A linear weighted method is used to convert the multi-objective problem into a single-objective one. An improved Harris Hawks optimization, where a fuzzy elitism strategy and a nonlinear escape energy updating strategy are introduced, is designed to avoid local optimum solution, and improve the quality of solution. The data from Line S1 railway of the City of Taizhou and its connecting high-speed railways is taken as the case study. Study results show that the value of the objective function of the model decreases by 36.7%, compared with its initial solution. Compared with the benchmark Harris Hawks optimization, the train operation cost, the total generalized travel cost, and generalized objective of the model decreases by 8.09%, 7.04%, and 7.56%, respectively. These results prove that the designed algorithm is suitable and has a higher accuracy. Within the proposed model, train capacity and efficiency of passenger transfer in different time periods have been taken into account in the model, which is helpful to improve the coordination among different railway lines under the condition of multi-network integration.

Pathway for Integrated Development of Port and Clean Energy Under Strategy of Carbon Peaking and Carbon Neutralization in China

JIANG Yipeng, YUAN Chengqing, YUAN Yupeng, DONG Mingwang, JIANG Tao, ZHONG Xiaohui, TONG Liang

2023, 41(2): 139-146. doi: 10.3963/j.jssn.1674-4861.2023.02.015

Abstract(944) HTML (324) PDF(56)

Abstract:
Clean energy provides a viable approach towards environmentally friendly, low-carbon development within the port industry. However, the lack of comprehensive planning and efficient utilization of accessible clean energy sources hinders a sustainable growth of China's port industry. To achieve the"Carbon Peaking and Carbon Neutrality"goals for ports and address challenges of high energy consumption and carbon emissions, the limitations within the integration process between the Port and Clean energy are investigated both domestically and abroad. Key issues are identified, such as limited types of clean energy, low penetration rates of clean energy, challenges in multi-energy grid technology, and the absence of integrated green hydrogen applications. This paper studies the endowment of natural resources in ports and analyzes their energy consumption forms. In addition, a comprehensive energy system for ports is proposed, which involves multi-energy integration across three aspects of"source-grid-load", forming an architecture that focuses on reducing carbon emissions and enhancing self-sufficiency in energy. Furthermore, the structure of energy integration system, which includes power generation and hydrogen consumption, is further refined. This system identifies a topology structure for the port's comprehensive energy system, centered primarily on power systems, electrical grid, and energy consumption of the terminal. Developmental paths are also outlined, including the policy support, technological innovation, and demonstration trials. Moreover, taking Chuanshan port region of Ningbo Zhoushan Port as an example, this paper devises a waterborne port-vessel multi-energy integration application system architecture, with core components spanning energy, control, grid, and load layers. By utilizing the natural resources of the port region, a sustainable energy provision architecture has been devised, which is anchored by wind power and supported by supplementary clean energy options. The power generation replaced by clean energy is anticipated to surpass 14 MW, with an associated carbon emissions reduction exceeding 20 000 t. This paper exemplifies the integration of ports with clean energy, substantiating its practicability and efficacy, as well as providing a scientific reference for the future ecologically conscious and low-carbon development of port industry in China.

Modeling and Analysis of the Effects of Government Incentives onto Reduction of Ship Carbon Emission Based on an Improved Principal-agent Model

LIU Yi, BU Xinru

2023, 41(2): 147-156. doi: 10.3963/j.jssn.1674-4861.2023.02.016

Abstract(576) HTML (231) PDF(17)

Abstract:
Currently, the incentives for carbon emission reduction mainly focus on carbon pricing, carbon taxes, and reduction of ship speed, while there is no subsidy incentives for shipping companies implementing carbon reduction measures. Due to the differences in operational modes, emission reduction strategies, and asset compositions between shipping companies and ordinary enterprises, the assumptions of traditional principal-agent model fail to meet the need to analyze the effect of government incentive onto the shipping industry. Therefore, specific improvements to the existing principalagent models are required. The purpose is to achieve the maximum social benefit with limited government funds, while the shipping companies that make best efforts will receive maximum profits. Considering the exogenous uncertainty of the effect of policies onto carbon reduction for risk-averse shipping companies, a government incentive model for the reduction of ship carbon emission is proposed. Since the government cannot always fully observe the efforts of a shipping company for reducing emissions, the scenarios with complete and incomplete information are independently analyzed. Setting the optimization objectives separately for the government and the shipping companies to maximize their own benefit, the optimal reward-penalty coefficient for the government and the optimal level of efforts for reducing carbon emission from the shipping companies are found. The optimal incentive contract is also studied for the government to regulate the shipping companies. The parameters for internal and external factors determining the efforts for reducing carbon emission by the shipping companies are discussed, as well as their impacts on the optimal incentives provided by the government and the relevant parameters. Study results show that the optimal incentive coefficient decreases at a decreasing rate when the variance of exogenous random variable and the absolute risk aversion coefficient increase. When the variance is 8 and the risk aversion coefficient is 4, the rate of decrease stabilizes. At this point, the shipping companies exhibit a strong aversion to the risks of reducing carbon emission and a strong resistance to implementing the measures. The cost coefficient and the coefficient influencing the level of efforts for reducing emissions jointly affect the intensity of the companies to implement measures. When the level of efforts is high, the government incentives initially increase quickly and then level off as the cost for the shipping companies increase. This implies that the government aims to meet the needs of reducing carbon emission of the shipping companies within its limited funds to reduce difficulties within the implementation. However, once the incentives reach a certain level, further increase in incentive will not directly affect the motivation of shipping companies to reduce emissions. Due to the joint impact of the cost coefficient and the coefficient of the level of effort for emission reduction on the government incentives, there will be one optimal incentive under different shipping market conditions. Considering the development of shipping industry along the Yangtze River, when the cost coefficient is 0.5 and the coefficient of the level of effort for emission reduction is 3, the optimal impact of government incentive is achieved.

State-of-the-art and Prospect of Technical Standards for the Ships Powered by Hydrogen Fuel Cells

WANG Dongxing, WANG Zhe, ZHAO Fan, HAN Fenghui, JI Yulong

2023, 41(2): 157-167. doi: 10.3963/j.jssn.1674-4861.2023.02.017

Abstract(862) HTML (283) PDF(82)

Abstract:
With the increasing global concerns over energy and environmental issues, energy conservation and emission reduction have become a consensus among countries. There is an urgent need for the global shipping industry to stick to a green, low-carbon/zero-carbon development path, due to its considerable energy consumption and emissions. The development of ships powered by hydrogen fuel cells provides an opportunity for the shipping industry to achieve relevant targets for energy conservation and emission reduction. There are active research, promotions and demonstrations of ships powered by hydrogen fuel cells in China and many other countries, among which technical standards are an important foundation for the development of this particular technology. However, on one hand, the relevant domestic and foreign hydrogen energy technologies for marine applications and regulatory frameworks have yet to be developed. The corresponding standard systems are also incomplete with the lack of core standards. On the other hand, the standards from the automotive hydrogen energy industry, which are relatively more comprehensive, are not applicable to the maritime system, due to the inherent technical differences between the automotive and maritime industries. Therefore, based on the technological roadmap and CHAIN OF INDUSTRIALIZATION of ships powered by hydrogen fuel cells, this study systematically reviews the technical structure and the status quo of domestic and foreign projects and analyzes technical regulations of the hydrogen fuel and hydrogen fuel cell from Classification Society. The domestic and foreign hydrogen standards and regulations are also studied, classified, and systematically analyzed. Furthermore, a green shipping framework is developed, which mainly focuses on the application of ships powered by hydrogen fuel cells and covers the production chain of hydrogen production, storage, transportation, and refueling. Within the framework of the proposed green shipping system, technologies of vehicles and ships powered by hydrogen fuel cells are compared. Study results show that there are differences in the application settings, power, hydrogen storage, and refueling facilities, as well as significant similarities and differences in service life, starting conditions, space requirements, installation methods, ventilation, and airtightness requirements. Based on these findings, this study focuses on developing a standard system for ships powered by hydrogen fuel cells, referencing the applicability of hydrogen energy standards for road vehicles, developing the core standards for the chain of industrialization of ships powered by hydrogen fuel cells, and formulating safety regulations and standards for such ships. The paper aims to develop theframework and directions for technical specifications, certification, and product standards for the industry of ships powered by hydrogen fuel cells, specifically tailored to the green shipping system.

A Study on the Status Quo and Trend of Green Energy Technology for Shipping Industry

CHEN Gong, ZHU Yu, HAN Bing

2023, 41(2): 168-178. doi: 10.3963/j.jssn.1674-4861.2023.02.018

Abstract(889) HTML (378) PDF(61)

Abstract:
In the context of developing a sustainable maritime transportation system, the green energy technology is studied from the perspective of the domestic and foreign carbon emission policies and the discharge requirements of Arctic waterway commercial navigation. In order to develop a shipping industry with a low-carbon emission, it has become a major task for the relevant researchers to study the green energy technology. Focusing on the innovative study of energy technology, the application potential of green fuels to achieve carbon neutrality in shipping is first studied. The applications and research progress of hydrogen, ammonia, and green methanol on ships at home and abroad are reviewed. A preliminary evaluation of green energy for shipping industry is conducted from the perspectives of the Technology Readiness Level and the Commercial Readiness Index. The main problems of using green energy technology in the shipping industry are analyzed and discussed from the following five aspects: fuel production, fuel infrastructure, power system, ship design, and navigation operation. Study results show that the main factors restricting the application of green energy technology are the high cost of fuel, the difficulty of large-scale fuel supply, the lack of basic supporting facilities and the immature key equipment technology. Compared with traditional heavy diesel and LNG fuel technologies, there is no single fuel that has comprehensive and overwhelming technical advantages to completely replace existing traditional fuels. However, based on the status quo and development trends, it can be predicted that methanol will become the main energy to achieve phased carbon reduction. Hydrogen and ammonia are more advantageous in achieving carbon neutrality and are suitable for achieving carbon neutrality of inland and marine vessels, respectively. Based on the issues identified from the existing studies, the suggestions are offered from the following three aspects: supply chains, ship technologies and standards.

2023, 41(2): 179-184.

Abstract(344) HTML (393) PDF(17)

Abstract:

2023 Vol. 41, No. 2