Green Connected Automated Transportation and Safety

Table of Contents

1. Trends Analysis of Domestic Waterway Traffic Safety Production Based on Regional Characteristics in China

   Guobo Wang, Guanquan Chu, Chao Han

   Abstract

   In order to improve the level of waterway traffic safety production in China, based on the significant differences in waterway traffic safety production in different regions, analogy and trend analysis methods are used to analyze the safety production situation of water transportation in different regions from multiple perspectives such as shipping, ports, ships, and accidents. The results show that the waterway freight volume in different areas maintains a relatively stable growth, and the long-distance transportation of bulk cargo is the main development trend. The throughput growth of inland river ports slowed down, and the throughput of coastal ports continued to increase. The average deadweight of transport ships in inland water shows a significant linear growth trend, while the passenger capacity of ships shows a fluctuating downward trend. The distresses are mainly concentrated in coastal waters, and the characteristics of the distresses in different areas are different. The number of distress in coastal waters is mainly related to the number of severe weather, while the number of distress in Inland River has the highest correlation coefficient with the number of ships.

2. Study on Route Optimization of Seasonal Hot Spots

   Haixia Zhao, Ting Tong, Jiangping You

   Abstract

   With the adjustment of holiday policy and the rapid growth of consumers’ demand for tourism, severe overcrowding in hot senic spots has become one of the main bottlenecks that perplex the development of many scenic spots in China. Based on the relevant literature and relevant data obtained from practical investigation, the factors that affect the path planning of hotspots are reasonably weighted by entropy method, and short-term prediction method is used to predict the passenger flow. Based on the improved A* path search algorithm and careful consideration in the various costs of different scenic spots in the scenic area, we can determine which scenic spot should come next, and finally get the optimal path. When tourists arrive at different scenic spots, they will make further progress according to the overall situation of the scenic area and update the optimal path. This paper takes Huangguoshu Tianxing Bridge scenic spot as an example to verify the improved A* path search algorithm. When taking Tianxing Bridge as the starting point, the optimal path is obtained by predicting the passenger flow of each scenic spot in the scenic area and comprehensively evaluating the influence of various factors, taking into account the total cost of each scenic spot. This paper considers the single way for tourists to visit—walking. In fact, for the problems of scattered scenic spots and considerable distance between them, sightseeing cars and shared bicycles would be arranged in the scenic spots. But by changing the value of traffic speed, the arrival time of each scenic spot can be calculated. The result has wide applicability, and the optimal path planning method has certain flexibility, which can be widely used in the route planning of other scenic spots.

3. Research on Energy-Saving Driving for Transport Vehicles Considering Actual Load

   Chunming Li, Tao Zhang, Xiaoxia Sun, Chunming Shao, Guozhu Wang

   Abstract

   To improve the economy performance of plug-in hybrid electric transport vehicle (PHETV) on urban roads, a pseudo-spectral based speed planning method is proposed for connected vehicles to assist driver to pass continuous intersections with fewer stop-and-go behaviors. Firstly, the effective green light time is estimated based on the traffic flow obtained by V2I. Secondly, the actual vehicle load is identified through neural network which is trained by the historical data. Finally, the time-domain optimal control problem with piecewise head and tail constraints is constructed and solved by the pseudo-spectral method to minimize the travelling energy consumption. Simulation results show that the proposed control strategy can make fuel consumption of PHETV reduce by 6.71% compared with traditional method.

4. Research on the Coupling Relationship Between Passengers and Vehicles Based on Anylogic Simulation

   Jiahui Li, Siyuan Hao, Chengwu Jiao, Nale Zhao, Keman Wu

   Abstract

   With development of the urbanization process, more and more integrated passenger transport hubs were constructed. Some softwares are used to simulate traffic flow around the hub to plan traffic form layout and simulate the passengers in the hub to optimize facilities layout and walk access. In fact, passenger transport hub is the connection between passengers and vehicles, vehicles pick-up passengers in hubs, and passengers can transfer different means of transportation in hubs. But there are few simulation researches on the coupling relationship between passengers and vehicles. In fact, the poor connection between pedestrians and vehicles will affect the hub operating efficiency. Therefore, this research was based on the taxi picking up passengers in waiting area, explored the coupling relationship between passengers and vehicles, and realized the key simulation technology that one vehicle leaves with passengers then the vehicle behind it pulls into the empty parking space. This technology will increase the accuracy of hub simulation and improve 3D demonstration effect.

5. Coevolution Dynamics Model of Urban Multimodal Transportation System Under Limited Exploitable Land Resources

   Rong Li, Weili Kong, Liwen Zhang, Peng Shu

   Abstract

   Urban multimodal transportation system evolution is a complex dynamical process under limited exploitable land resources. In this paper, a five-level coevolution dynamics model among land development, population, roads, private vehicle and bus ownership is developed to capture the growth mechanisms of urban traffic systems. It is convergent under certain conditions through stability analysis. This model can capture the observed co-evolution law and predict the evolution trend of urban multimodal transportation system in the future by taking Beijing city as case study. It is of great significance to urban traffic planning and management.

6. Traffic Congestion Improvement Scheme of the Corridor Between Beijing Deputy Center and Central Urban Area

   Chuanjiao Sun, Panyi Wei

   Abstract

   In order to solve the congestion problem between Beijing deputy Center and Central Urban Area after the movement of Beijing government, the main corridors between the two areas are studied on the main congestion index and the importance of the area. A model about area affections are constructed by the congestion and the importance. Based on the model, the main congestion locations are given and the improvement measures are listed.

7. Research on the Plan of Container Train Operation in Sea-Rail Intermodal Transportation Under the Uncertain Demand

   Zhiyi Lei, Zhikuan Ma, Jéssica Saturnino Sabino de Sousa, Yifan Zhu

   Abstract

   The sea-rail intermodal container train is an important part of the sea-rail intermodal system. Its operation plan directly affects the detention time of sea rail intermodal containers in the port. In view of the characteristics of the large fluctuation of the container flow to the port, the Gauss hidden Markov model is employed to predict the container flow and a multi-objective mixed integer-programming model is established to minimize the detention cost of container and the cost of the trains. As an example, different scenario needs of Yangshan Port in Shanghai are forecasted through Gauss hidden Markov model, and the container train operation scheme under different scenarios is optimized for the terminal front loading and unloading container scheduling scheme.

8. Research on Setting of Warning Signs for Highway Tunnels in High Altitude Areas

   Jian-you Zhao, Wei-guo Xie, Wan-li Han, Wan-jiang Guo

   Abstract

   In order to study the driving conditions of drivers on highway and tunnel sections in high altitude areas, took some sections of Gong Yu highway as the research object, and the driving speeds of the subjects were collected on sections at different altitudes to verify the influence of road environment on drivers’ behavior characteristics in high altitude area. Aimed at the problem of irregular installation of traffic facilities on highway and tunnel sections in high altitude areas, and considered the distribution and species of animals along highway in high altitude areas, measures have been taken to install animal crossing warning signs. Based on the visual recognition characteristics of drivers, construct a model of the warning signs pre-distance for highway tunnels in high altitude areas, and get the speed limit signs pre-distance reference table.

9. Analysis on the Road Capacity: Connected and Automated Vehicle Platoons in the Mixed Traffic Flow

   Ke Ma, Hao Wang

   Abstract

   The intelligence of connected and automated vehicles (CAV) are not enough to deal with the complex road environment. CAVs based on platoon can reduce external interference and further improve road capacity. Therefore, a mixed traffic flow formula considering ACC, CACC system and leading CAV information interaction based on the CAV platoon is constructed. By analyzing the four main parameters (the size of platoon, ACC headway gain, CACC headway gain and leading CAV communication headway gain), we find the most important platoon parameters that affect the road capacity.

10. Real-time Trajectory Prediction of Unmanned Aircraft Vehicles Based on Gated Recurrent Unit

    Rong Tang, Zhao Yang, Jiahuan Lu, Hao Liu, Honghai Zhang

    Abstract

    In view of the increasing number of unmanned aircraft vehicles (UAVs) in low altitude airspace, which leads to the increasing difficulty of surveillance, a real-time trajectory prediction method based on the gated recurrent unit (GRU) is proposed. The parameters of the GRU model are determined including the number of hidden layers, the cell sizes of a single hidden layer and the time step. Historical trajectory data are used to train and test the model. The result shows that the mean absolute errors (MAE) of altitude, longitude, and latitude predicted by the GRU model are 0.86 m, 2.85 m, and 3.65 m, respectively. The prediction results of the GRU model are compared with the long and short-term memory (LSTM), the support vector regression (SVR) and the autoregressive integrated moving average model (ARIMA). It is found that the GRU and LSTM outperform the two other models. The LSTM prediction accuracy is slightly higher than that of the GRU model. However, the space complexity of the LSTM model is much larger than that of the GRU model, and its training time is much longer.

11. The Capacity Model of Urban Multimode Transportation Network Based on Conbined Travel Mode

    Guozhu Cheng, Chuanmiao Zhou

    Abstract

    In order to research the carrying capacity of the urban multimode transportation network to for the number of trips, this paper constructs the corresponding super network and travel cost functions based on combined travel modes. For a given current O-D pattern, a bi-level programming model is formulated to estimate the reserve capacity of the multimode transportation network, the upper-level model aims at maximizing the amount of travel, the lower-level model describes the traveler’s behavior, conforms to the stochastic equilibrium distribution. Finally, considering the limitation of transfer time and level of service, a heuristic method based on incremental assignment technique is proposed to solve this model.

12. Autonomous Train Dispatch Method for Regional Rail Transit Based on Deep Reinforcement Learning

    Wei Wei, Ling Liu, Jun Liu, Bo Zhang

    Abstract

    Train dispatching of rail transit is an important issue for implementing the transportation plan and ensuring the orderly operation. Recognizing that existing centralized dispatch methods lack efficiency and flexibility, a novel concept of autonomous train dispatch of rail transit is proposed. Additionally, the Deep Reinforcement Learning (DRL) is incorporated into the autonomous train dispatch (ATD) model, and the resultant DRL-ATD model considers both dispatch and control objectives. The Double DQN (DDQN) algorithm is used for the model training, and a prioritized experience replay mechanism is further introduced to improve the stability and convergence. Using a case study involving the Chongqing Rail Transit in China, it is found that the DRL-ATD model can ensure the safety and punctuality of rail transit trains; it also reduces energy consumption by 1.62% during low passenger volume periods and passenger waiting time by 6.83% during high passenger volume periods, compared with the existing centralized method. Sensitivity analysis of the dispatch policy shows that the DRL-ATD model can effectively balance energy consumption of train operations and the total waiting time of passengers by adjusting the train operation schedules automatically, thereby providing the advantages of flexibility and dynamic adaptability for train operation.

13. Research on Energy Delivery of a Series–Parallel Hybrid Electric Vehicle Under Different Driving Conditions

    Chunming Li, Xiaoxia Sun, Chunming Shao, Lining Yang, Chenglong Shu, Danhua Niu, Guozhu Wang

    Abstract

    In series–parallel hybrid electric vehicle, the relationship between power flow and heat flow is very complex. The dynamic thermal characteristic of series–parallel hybrid electric vehicle is closely related to vehicle energy management strategy, external environment and driving conditions. In this paper, the matching of the vehicle working mode with the vehicle operating conditions is studied. The theoretic analysis of vehicle energy distribution under different typical operating condition is conducted. A coupled whole vehicle simulation model which is built in MATLAB/SIMULINK and GT suite is proposed. According to the theoretical qualitative analysis and simulation quantitative analysis, the corresponding relationship combined heat response and power consumption between power flow and heat flow is acquired. Study results show that the coupled simulation model can be effectively used in the research combined dynamic heat response, the power and heat delivery and power consumption under different operating conditions and working modes. This integrated research method combined theoretical qualitative analysis and simulation quantitative analysis is an efficient way in the multi-disciplined complex vehicle system energy management study.

14. Information Extracting and Analysis on Safe Operation of Online Car-hailing Based on Text Mining: The Case of China

    Si Yang Liu, Jian Rong, Chen Jing Zhou, Yi Wang, Zi He Lin

    Abstract

    The recent rise of online car-hailing has successfully promoted the booming of sharing economy in transportation and sets up new mode for intra-city trips. However, it comes with serious risk for passengers, drivers and online car-hailing platforms (OCPs). Few researches have explored the risk of OCPs services despite most research focus on the safe assessment/evaluation system. As a special transportation service enterprise, the research on the safe of OCPs is obviously insufficient. This paper explores the focus of online car Hailing industry-related researches, and build a framework for OCPs services by identifying 8 topics from 883 research papers. Natural language processing method are adopted to achieve the target and specifically, the TF-IDF algorithm and LDA model are applied in the study. The results show that regional policy background provides basic safe situation for OCPs services, and stress the importance of safety improvement from both macro vision and detailed optimization method. Besides, current disputes and dilemma of OCPs safe operation and the legal status and responsibility of OCPs should be paid more attention. In addition, shielded from various risks (e.g., functional risk, time risk physical risk, financial risk) is a key to get the advantage in the competition for market share. Finally, practice and planning of OCPs services in regions is needed to promote the safe operation of OCPs services.

15. The Method of Division and Classification Quantification for Snow and Ice Removal on Airport Pavement

    Ya-ping Zhang, Yu Zhang, Xiao-qing Xing, Ye Chen, Nan Yang, Jian Mao

    Abstract

    In the northern cities of China, the winter snow and ice on the airport pavement seriously affects the airport operation efficiency. It is necessary to study the area division and classification quantification of airport pavement snow and ice removal operation in order to scientifically dispatch the equipment group of airport pavement for snow clearance and de-icing. Based on the analysis of the importance of different airport pavements, the airport pavement is divided into different snow removal areas, and the initial priority quantification model is proposed. The gray difference network method is used to predict the operation time of the airport pavement snow removal. Then, the dynamic priority model of the airport pavement is established with the initial priority as the input, and the dynamic status of the airport pavement priority in each stage is generated, and output dynamic priority matrix. Finally, the process of snow and ice removal in a large hub airport is simulated, and the simulation results show that the proposed method is feasible.

16. Research on Object Detection and Location Using Semantic Segmentation

    Huijie Zhu, Yan Cai, Yingchun Guo, Xiang Feng, Wenya Zhang

    Abstract

    Object detection and location generally require multiple sensors and multiple deep networks to achieve. For a specific application scenario, this paper proposes an algorithm based on semantic segmentation to display object detection and location at the same time. Due to the small size of the target in the image, especially the vertex of the target occupies very few pixels, the HRnet semantic segmentation algorithm with higher resolution is adopted. Due to the unbalanced class distribution and the small number of training samples, data augmentation and Lovasz Softmax loss function are used in a targeted manner. After experimental verification, in this data set, HRnet is superior to other algorithms. And by using data augmentation and Lovasz Softmax loss function, the accuracy and IoU have been greatly improved, especially the target vertex area with few samples.

17. General Standard Model for Safety Risk Assessment of Highway Bridge Construction

    Yuqian Wang, Dianliang Xiao

    Abstract

    With the high-level development of highway construction in China, it is very important to carry out standardized and universal risk assessment of bridge construction, and to find out the high-risk operations and risk event type distribution of the supervised bridges, by which is the basis for the related government departments to conduct focused risk supervision. This paper establishes general standard model for safety risk assessment of bridge construction in response to the departments’ requirements of generality, standardized, convenient, easy to promote, and as objective as possible, as well as the limitations of current guideline “Safety Risk Assessment Guidelines for Highway Bridge and Tunnel Engineering Construction”. This model is applicable to various risk events in different process for all types of bridges, including effective, reliable and measurable 14 first-level indexes and 82 second-level indexes based on 4M factors, which realized modularization and standardization of possibility assessment index system. After being verified by the application of Nansha bridge (the 2nd Humen Bridge) and many other bridges, the model is also used in the revision of “Safety Risk Assessment Guidelines for Highway Bridge and Tunnel Engineering Construction”.

18. Modeling Simulation and Stability Evaluation Analysis of Double-Trailer Combination

    Yue Li, Hao Zhang, Hongwei Zhang, Chengqiang Zong, Chuanjin Ou, Xueli Zhang

    Abstract

    In order to analyze the handling stability of double-trailer combination, the TruckSim simulation model was established, and the whole vehicle road test was carried out. The comparison between the road test results and the simulation results verifies the rationality and accuracy of the simulation model, indicating that the simulation model can reflect the characteristics of the double-trailer combination. By setting the steering wheel step input and the single line change test, the steering stability of the double-trailer combination was analyzed, concluding that Dolly is an important factor affecting the double-trailer combination. The simulation analysis results can provide the direction of improvement and technical guidance for the driving stability of double-trailer combinations.