Trajectory prediction is a critical technique for enhancing the efficiency of marine traffic control and maintaining the vessel navigation safety. In general, accurate trajectory prediction is examined in both temporal and spatial dimensions, including dynamic motion pattern modeling and capturing spatial interactions between vessels. Existing approaches have made some headway on the aforementioned concerns, but they tend to disregard the dynamic interactions between vessels and only evaluate static properties. To address this challenge, this paper proposes a novel Gated Spatio-Temporal Graph Aggregation Network (G-STGAN) for vessel trajectory prediction. Specifically, Spatial Gated Encoder (SGE), a variant of graph convolutional networks, is presented to describe the spatial interactions between vessels. In the temporal dimension, we designed a temporal gated encoder (TGE) to effectively fuse short-term and long-term temporal dependencies. Furthermore, the spatial and temporal features from the SGE and TGE modules, respectively, are aggregated by using a temporal convolutional network (TCN) to perform downstream prediction tasks. Experiments on three real-world AIS datasets demonstrate that G-STGAN can achieve competitive prediction performance in terms of accuracy and robustness.
Aiming at the drawbacks of traditional artificial potential field method in local path planning for USV (unmanned surface vehicle), an improved artificial potential field method is proposed. Firstly, establish a model for ship collision risk based on the quaternion ship domain. In the gravitational potential field model, Add the gravity of the given route. In the repulsion field model, the influence factors such as relative position and relative velocity are introduced. Four subdivided zones around the own ship are defined to determine the corresponding virtual repulsive potential to ensure International Regulations for Preventing Collisions at Sea (COLREGS)-constrained behaviour for the own ship’s CA actions. The rationality and superiority of the proposed improved artificial potential field method are verified by simulation and comparison experiments Simulations demonstrate that the method is fast, effective, and deterministic for path planning in complex situations with moving target ships, and has been proved to work effectively in a complex navigation environment.
The lane-changing maneuvers of vehicles at freeway merging areas can result in a substantial increase in crash risk. The implementation of a connected and autonomous vehicle (CAV) environment is anticipated to decrease the associated conflict risk of such maneuvers, by utilization of surrounding traffic information. To investigate the potential conflict situations that may arise during multi-vehicle interactions by taking advantage of CAV environment, this study proposes a Bayesian hierarchical algorithm to identify traffic conflicts in real-time during mandatory drivers’ merging behavior. The Bayesian hierarchical model consists of observation vehicle conflict layer and prior distribution layer. The MCMC sampling method was utilized to calibrate the model parameters by generating a Markov chain. Two methods including image discrimination and posterior distribution comparison are combined to ensure reliable convergence diagnosis of Markov chain. The results show that the Markov chain for each parameter gradually stabilizes over iterations and the posterior probability density is consistent with the specified prior function and likelihood function. Additionally, the flexibility of this approach significantly enhances the ability to analyze traffic conflict from the real world using applied statistics. The findings of this research have the potential to encourage cautious driving behaviors and provide useful insights for driver safety in the future, especially in the context of CAVs.
Maritime transport plays an increasingly important role in global shipping. In the meantime, with the development of the shipping industry, the frequency of maritime traffic accidents is also increasing. More importantly, the increasing number of maritime traffic accidents with serious consequences in the past 20 years cannot be ignored. Analysis of major accidents can help identify trends and prevent future accidents. The data set of maritime traffic accidents from 2002 to 2022 is analyzed in this study, focusing on the ship size, ship age, ship type, and other factors of ships involved in major accidents. Regarding to the accidents that with serious consequences in the past 20 years, Bayesian rules and least square method are synthesized to model the historical accident data. This method compares the accident risks of different sizes and types of ships in pairs, calculates and analyzes the impact of each risk factor on the occurrence of serious accidents. The results indicate that the risk of major accidents on small and medium-sized ships is higher than that on large ships. It also indicates that improving the safety of auxiliary ships and bulk carriers is significant in preventing major accidents. The results are valuable for decision-makers in evaluating the level of navigation risk and reduce the frequency of future accidents.
Due to the harsh working environment of gearboxes and strong noise interference, the accuracy of fault identification is seriously affected. In addition, the traditional ResNet network has a large number of network parameters, resulting in slow training speed. Experiments show that the accuracy of the proposed Fast-SeResNet network is maintained at 99% compared to the traditional ResNet network, while the number of training parameters has been reduced by an order of magnitude, and the training time for each Batch is reduced from 66 seconds to 10 seconds with the same hardware support. The results show that the Fast-SeResNet network structure can improve the diagnostic speed to a large extent in a noisy environment with a small improvement in network accuracy, and has some practical value.
International dual carbon strategy and net zero action has significantly facilitated electric logistics vehicles. However, the fire accident is a threat for the logistic safety. A data-driven Bayesian network model is proposed to assess the fire consequence in electric logistics vehicles (ELV) based on 469 fire accidents, focusing on safety concerns, fire probability, and the evolution process. Firstly, accident data is collected to identify risk factors. Secondly, the EM algorithm is used to obtain the conditional probability table of the Bayesian network and establish an assessment model for fire consequence in ELV. Thirdly, FDS software is used to simulate specific fire scenarios in ELV and collect fire parameters. The results show that there is a 61.5% probability that the thermal runaway of the electric vehicle (EV) battery will lead to a complete vehicle fire, with a 67.1% probability of freight being lost totally. After the combustion becomes stable, the area 2.45m away from the ELV is relatively safe, and the thermal radiation intensity of the rear door of the vehicle compartment is lower. Therefore, targeted design and management recommendations should be proposed to reduce the risk of fire.
With the development of large-scale ships and the shipping industry, ship contact accidents also occur occasionally. In order to enhance navigational safety and control the risks from contact accidents, it is necessary to investigate the characteristics of global ship contact accidents and in-depth analysis of their causes. This paper focuses on the feature and cause analysis of contract accidents by incorporating the decision-making trial and evaluation laboratory (DEMATEL) and association rules (AR) based on global ship contact accidents from 1977 to 2022. First, the contact accident data are dealt with regarding involved ships, accident severity, and spatial and temporal distribution of contact accidents, and a frequency matrix of contact accidents is proposed. Second, a co-occurrence network model is proposed to identify the critical features of contact accidents. Third, the frequency matrix is further processed as the complex network’s input matrix using the DEMATEL method and AR. Furthermore, the complex network of influencing factors of ship contact accidents is used to identify the relevant factors affecting the severity of ship contact accidents and pollution degree. Then, the centrality of the complex network is analyzed to obtain the critical factors affecting the severity of ship contact accidents and pollution degree. The research results can provide theoretical support for risk management of contact accidents.
To address the potential impacts of non-steady-state heat source conditions on the operating parameters of the supercritical carbon dioxide (SCO 2 ) Brayton cycle power generation system in ship power systems, this study establishes a dynamic numerical simulation model of a simplified SCO2 Brayton cycle power generation system using the MATLAB/SIMULINK platform. Based on this model, transient operation characteristics of the system are analyzed. The variations in operating parameters of the thermal cycle system are simulated when flue gas parameters change. The effects of heat source temperature fluctuations on the inlet and outlet parameters of system components and the net power output are analyzed, along with control strategies. The results indicate that during heat source fluctuations, the $CO_{2}$ may operate in a transcritical state, and the system pressure may exceed its maximum tolerable value. After implementing control systems, it is found that when the flow rate remains constant within the system, it is preferable to allow the turbine inlet temperature to vary based on waste heat source fluctuations. The combination of low-temperature and inventory control shows good performance in regulating the system’s minimum temperature and pressure, enabling the system to operate at its optimal design points.
As the backbone of the metropolis, the metro is developing towards networking and complexity. Random and unpredictable emergencies will significantly impact the metro and transportation network. At present, the emergency bridging bus is an important strategy to ensure the reliability of the metro system, and the depots are the primary source of the bridge bus. Whereas, the previous research on the location of the depots does not consider the randomness of emergencies and the reliability of the new transportation system after adopting the bridging strategy. To fill the research gap, we propose a multi-objective optimization model for depot location based on metro vulnerability and information entropy evaluation theory. First, according to the topological structure of the metro, passenger demands and other factors to obtain the importance of network nodes. Secondly, the information entropy evaluation method is innovatively proposed to evaluate the network performance after the location of the depots from three perspectives: the network adjacency attribute, the ability of the depots to transport the bridging bus to the interrupted interval, and the rescue ability of the bridging buses dispatched by the depots. Finally, a multi-objective optimization model is established with optimal network performance and minimum depots construction cost. The results show that the model proposed in this paper can provide a better bridging effect and reduce passenger delay while effectively reducing construction costs. The study can notedly provide suitable and operable location strategies for urban traffic managers.
Due to differences in resource endowment around the world, domestic supply often cannot meet demand and relies on cross-border supply chains. To enhance the resilience of cross-border supply chains, it is necessary to increase independent controllability. This requires joint participation by both the government and enterprises. This paper constructed an evolutionary game model to study the path of government policy making and enterprise integration strategy selection in cross-border supply chains, and further explores the equilibrium conditions of enterprise supply chain integration behavior under government leadership. Numerical simulation analysis shows that under most parameter conditions, the government tends towards adopting an active strategy to improve supply chain controllability based on considerations of public interest and supply chain security. Enterprise’s decision-making willingness is affected by the relationship between supply chain integration costs and benefits.
Along with the increasing importance of maritime transportation, the supervision of vessel safety is also of great importance in port management. Port authorities in coastal countries have always been committed themselves to improve the efficiency of the supervision on vessel safety, thus reducing the illegal action of ship owners and ensuring the navigational safety in maritime transportation. In this research, a trilateral evolutionary game model among different stakeholders (i.e., port authorities, ship owners, social groups) in vessel safety supervision is established, aiming at exploring the optimal strategies of different stakeholders, as well as the key factors influencing the evolution of the supervision system. Replicator dynamic equations are produced to reveal the equilibrium points in the game model, and MATLAB software is utilized to conduct the numerical simulation experiments on these equilibrium points. As a result, ‘port authorities choose strict supervision, ship owners choose standard management and social groups choose participation in vessel supervision’ is demonstrated to be the optimal supervision mode, and the reward and punishment mechanism proposed by port authorities is proved to be the crucial factor affecting the decisions of different stakeholders. Useful suggestions and insights are provided to assist port authorities in better supervising vessel safety conditions, improving the overall vessel quality, and ensuring the maritime safety.
Lattice structures combined with light weight and better mechanical properties are widely used in the marine and aerospace industries. Therefore, the triply periodic minimal surface (TPMS) lattice structure is chosen as the object of study in this paper. Since there are few studies on the vibration of such structures, the vibration-related properties were investigated by designing positive and negative gradient volume fraction TPMS structures in conjunction with the gradient concept. The sweep and vibration tests show that the designed IWP-type structure can provide vibration isolation in the low frequency range. At resonance the positive gradient structure has a higher resonant frequency due to its higher stiffness, but the negative gradient structure has a higher damping ratio and a wide frequency isolation band gap.
Automatic vehicles are likely to operate in mixed traffic conditions, affecting cyclists’ existing behavioral decisions. However, the intention and decision-making mechanism of cyclists’ risk-taking behavior on shared roads are not clearly understood. The theory of planned behavior(TPB) has been proven to be effective in the study of travel behavior. Based on the theory of TPB, the Extended Theory of Planned Behavior (ETPB) research framework of risk-taking behavior intention was constructed by introducing safety perception, psychological motivation and technology cognition variables. Safety perception variables took into account the public’s perception of safety in the shared road environment with AVs, while psychological motivation mainly took into account riders’ past riding habits, risk perception bias and riding emotion factors. Secondly, questionnaire survey was used to obtain riders’ personal basic information and multi-situation risk-taking behavior willingness data, and corresponding structural equation model was constructed to explain the influence of variables on willingness. Finally, clustering people by means of K-means clustering algorithm and analyzing the psychological and behavioral characteristics of all kinds of people, so as to facilitate the selection of traffic management strategies. The results show that the extended TPB model can better explain the risk-taking behaviors of cyclists on the self-driving shared road. The influence of attitude and safety perception on the willingness to take risks is more prominent, and technology cognition plays an important indirect role. This study expands the research on the psychology of cyclists’ risk-taking behavior, which could help to promote the integration of autonomous vehicles into transportation systems and better management strategies.
Throughout the years, an upward trend can be observed in the data regarding motor vehicle drivers in China, particularly in the number and proportion of female drivers. As an integral part of the road traffic system, female drivers possess unique physiological and psychological characteristics, driving patterns, and risk perception distinct from their male counterparts. Consequently, it becomes imperative to conduct specific research aimed at analyzing the driving behaviors of female drivers. Accordingly, this study centers on female drivers as the subject of investigation, developing a model that explores the decision-making process and inherent risks associated with their driving behavior. By examining the role of biases in risk perception, risk awareness, and attitudes towards driving safety, the study aims to shed light on the relationship between these factors and female drivers’ tendency to take risks while driving. Furthermore, a “Gender-based Driver Risk Perception Bias Scale” is designed alongside modifications to the existing “Driving Safety Attitude Scale,” “Driver Risk Perception Scale,” and “Risk-taking Driving Behavior Scale.” A survey was conducted among 274 vehicle drivers, and the data obtained were analyzed using AMOS to establish a structural equation model. The objective was to discern the underlying mechanisms through which risk perception biases impact risk-taking driving behavior specifically in female drivers. The findings of the study indicated the following: ① Significant disparities exist between male and female drivers concerning three aspects of risk cognitive bias: unfamiliarity with traffic information, overconfidence, and attribution bias; ② Risk cognitive bias exerts a positive influence on female drivers’ propensity for risk- traffic accidents of female drivers. Zhengtaking while driving; ③ Attitudes towards driving safety negatively influence risk-taking driving behavior; ④ Risk perception negatively impacts female drivers’ risk-taking behavior; ⑤ Risk perception partially mediates the relationship between risk perception bias and risky driving behavior. This study holds considerable implications for enhancing road traffic safety, improving the overall safety of the traffic system, and providing more tailored driving training for female drivers.
As the last line of defense for marine safety, Port State Control (PSC) plays an important role in improving marine safety and protecting the marine environment. In order to ensure the consistency and effectiveness of the inspection, this study uses a Bayesian Network to construct PSC examines evolutionary analysis models. Based on the data of PSC inspection in Paris MOU from 2018 to 2021, conduct research on the existing main ship types. The results show that for most ship types, the existing PSC inspection methods can guarantee the standardization of inspection, even in the face of public health emergencies; however, there are also cases of excessive attention to specific defects. Meanwhile, the sensitivity study of the model also found a significant correlation between RO performance, Inspection type and ship detention. The similar approach is also applicable to the evolution analysis of other MoU regions. In addition, the results have reference value for strengthening ship supervision and ensuring ship safety.
The vehicle FCW (forward collision warning) system is supposed to considerable reduce the traffic accidents caused by pedestrian crossing. The pedestrian path prediction method in FCW is studied, and the five-freedom-degree vehicle model is analyzed, the limit condition where whether the maximum barking acceleration can avoid collision effectively is researched. The vehicle coordinate is divided according to the psychological safety distance of pedestrians, and an anti-collision warning system is designed on basis of the pedestrian initial position, the movement speed and path. The test results show that the designed anti-collision warning system can effectively warn the driver before the collision accident occurs and avoid pedestrian traffic accidents caused by the negligence of the drivers.
Port State Control (PSC) is of positive significance for strengthening ship safety, reducing pollution, and urging the implementation of conventions, and plays an important role in the shipping industry. Detention, as an option of port state inspection, is used to promote the compliance of ships with standards, but for shipping companies, it means economic losses caused by delays in shipping schedules. This study aims to use the ensemble learning method to predict the possible defect types and detention possibilities of the main operating ship types based on the existing PSC inspection data. Based on using historical inspection records to extract characteristic indicators, the stacking algorithm is used to construct a ship detention prediction model, and the accuracy and effectiveness of the model are verified by comparing with the existing main prediction models. The results show that the model proposed in this study can effectively predict ship defects and detention; meanwhile, there are obvious differences in the types of defects that may exist in different ship types. This study has reference value for improving the efficiency of port state inspection and strengthening ship management.
Parking assistant system can not only replace the driver to park, but also help to reduce the occurrence of accidents during parking. The path planning and control under the typical parallel parking scene cited in the national standard GBT 41630-2022 have been studied. Through the analysis of vehicle motion parameters and the design of the path planning method, the safety and accuracy during the parking process have been validated. Meanwhile, the control strategy can be optimized according to the real-time target trajectory, which is utilized to reduce the parking position deviation and angle deviation. The simulation results show the path planning method can meet the requirement of parking, and the parking process performers excellent.
Ship-based carbon capture (SBCC) and storage technology is an effective way to realize the and emission reduction of ships and the greening of the shipping industry. In order to solve the problems of low decarbonization rate and high energy consumption issues of SBCC systems, a carbon capture model of LNG ship exhaust gas with a 3MW main engine was constructed based on ASPEN PLUS. Optimized basic process including absorber tower packing height, stripper diameter, and absorption liquid concentration process parameters were 9m, 0.74m, and 37.5%, respectively. The intercooling process was added to further improve the system decarbonization rate, and an orthogonal text was designed to explore the ideal installation parameter of intercooler. The intercool load, location, and temperature were 70%, 18, and 10°C, respectively. The advantages and disadvantages of the carbon capture capacity of the two technologies were analyzed from the aspects of solvent flow ratio and reflux ratio. It was proved that at the same processing conditions, the carbon capture amount of intercooling process was always higher than that of basic process. Adjusted solvent flow ratio intercooling process was 6.60% higher than the base process, solvent usage was reduced by 12.27%. The unit carbon capture energy consumption magnified only 0.201 MJ/kgCO 2 .
The carbon capture process model was established for an LNG ship, and the solution concentration and flow rate analysis selection under design conditions were carried out. The effect of solution flow rate on the carbon capture rate, energy and other key factors in the carbon capture system under off-design conditions was studied, and further research on the exhaust gas treatment mass flow and solution flow rate control was conducted under 75% MCR conditions. The results of the study showed that the energy of CO 2 regeneration and CO 2 removal rate increased significantly with the decrease of main engine load. The solution flow rate at the optimum operating point decreases approximately linearly with decreasing main engine load, while the optimal liquid-to-gas ratio (molar) does not tend to change with changing main engine load, so the solution flow rate can be controlled approximately linearly to achieve the optimum capture operating conditions when the main engine load changes. The appropriate exhaust gas treatment mass flow and solution flow rate can help to further reduce the regeneration energy based on the high CO 2 removal rate. Up to 82.8% CO 2 removal rate can be achieved at the regeneration energy of 9.477 GJ/T CO 2 under 75% MCR condition. This study can provide a reference for the control of carbon capture effect under off-design conditions of ships.
