Global route planning has garnered global scholarly attention as a crucial technology for ensuring the safe navigation of intelligent ships. The comprehensive influence of time-varying factors such as water depth, prohibited areas, navigational tracks, and traffic separation scheme (TSS) on ship navigation in coastal global route planning has not been fully considered in existing research, and the study of route planning method from the perspective of practical application is still needed. In this paper, a global route planning method based on human-like thinking for coastal sailing scenarios is proposed. Based on the historical route’s information, and taking into full consideration those time-varying factors, an abnormal waypoint detection and correction method is proposed to make the planned route conform to relevant regulations of coastal navigation and the common practices of seafarers as much as possible, and better meet the coastal navigation needs of unmanned ships. Taking the global route planning of “ZHIFEI”, China’s first autonomous navigation container ship, as an example, the validity and reliability of the proposed method are verified. Experimental findings demonstrate the efficacy of the proposed method in global route planning for coastal navigation ships. The method offers a solid theoretical foundation and technical guidance for global route planning research of unmanned ship.

Precise comprehension about the impacts of drivers' rational and perceptual characteristics on their behavioral decisions is crucial for the accurate prediction of driving behavior. In the previous research on driving behavior, drivers were regarded as homogeneous and absolutely rational individuals. To overcome this limitation, the coupling effects of bounded rational cognition and diverse emotions are considered, and a driving behavior model is proposed based on Dynamics Psychology. There are two parts in this research. In Study 1, the information entropy theory is applied to describe the vehicle cluster situation, and a method is established to quantify the cognitive uncertainty of the vehicle cluster situation for drivers in diverse emotions. In Study 2, with consideration of bounded rational and emotional cognition, the impacts of the vehicle cluster situation and drivers' features, which are their demands for the driving goals including safety, efficiency and comfort, emotional states, and cognitive characteristics, on driving behavior are uniformly expressed as the behavioral driving force, and a prediction model for driving behavior is proposed based on Dynamics Psychology. The results of validation based on the virtual driving data show the prediction accuracy of the proposed model for various driving behaviors of drivers in diverse emotions is over 80%. The results of verification based on NGSIM data suggest that the prediction accuracy of the proposed model for the natural driving behaviors is 82.07%. The research results can contribute to the study on the intrinsic mechanism of driving behavior and provide theoretical support for the development of traffic simulation, personalized active safety systems, human-machine interaction, and brain-inspired autonomous driving.

Local path planning, as an essential technology to ensure intelligent ships’ safe navigation, has attracted the attention of many scholars worldwide. In most existing studies, the impact of COLREGS has received limited consideration, and there is insufficient exploration of the method in complex waters with multiple interfering ships and static obstacles. Therefore, in this paper, a generation method for a time–space overlapping equivalent static obstacle line for ships in multi-ship encounter scenarios where both dynamic and static obstacles coexist is proposed. By dynamically inferring ships’ encounter situations and considering the requirements of COLREGS, the influence of interfering ships and static obstacles on the navigation of the target ship at different times in the near future is represented as static obstacle lines. These lines are then incorporated into the scene that the target ship encountered at the path planning moment. Subsequently, the existing path planning methods were extensively utilized to obtain the local path. Compared with many common path planning methods in random scenarios, the effectiveness and reliability of the method proposed are verified. It has been demonstrated by experimental results that the proposed method can offer a theoretical basis and technical support for the autonomous navigation of unmanned ships.

Accurate identification of driver’s distraction is of great significance to prevent traffic accidents. In this study, the information of driver’s body joint points is fully utilized, and a machine vision-based biomechanical distraction identification method is proposed. A real vehicle driving experiment is designed and organized. The normal driving data and biomechanical distraction data of 20 participants are collected to construct a data set. The Lightweight OpenPose network is trained to extract the position information of the driver’s joint points. A number of characteristic parameters such as the driver’s limb angles or Euclidean distance between the joint points can be calculated based on the position information. The experimental data are preprocessed using moving average filter and factor analysis. The biomechanical distraction identification model is trained with Particle Swarm Optimization (PSO) and Probabilistic Neural Network (PNN), and an accuracy rate of 91.5

Accurately estimating the real-time low-frequency motion state of the dynamic positioning (DP) ships under the time-varying environmental disturbances and effectively filtering out the environmental noise are crucial for enhancing the control accuracy of the DP system, ensuring safe ship operations, and minimizing energy transfer losses. With the principle of strong tracking, a strong tracking variational Bayesian unscented Kalman filter algorithm based on multiple fading factors is proposed and applied to the DP system. Firstly, the nonlinear system is unscented transfer, and the inverse Wishart distribution is adopted to model the measurement noise covariance matrix (MNCM). The variational Bayesian approach is used to recursively estimate the MNCM and system state vector. Secondly, to improve the correction precision of the predicted error covariance matrix (PECM), the multiple fading factors are introduced. In conjunction with the exponential weighting method of the fading memory, the estimated MNCM is considered as a time-varying parameter. Finally, the DP simulation model is constructed and the simulation results show that the proposed adaptive filter effectively improve the estimation accuracy and stability under the time-varying environmental perturbations. Furthermore, it significantly augments the adaptivity and robustness of the DP system.

Traffic lights detection and recognition (TLDR) is one of the necessary abilities of multi-type intelligent mobile platforms such as drones. Although previous TLDR methods have strong robustness in their recognition results, the feasibility of deployment of these methods is limited by their large model size and high requirements of computing power. In this paper, a novel lightweight TLDR method is proposed to improve its feasibility to be deployed on mobile platforms. The proposed method is a two-stage approach. In the detection stage, a novel lightweight YOLOv5s model is constructed to locate and extract the region of interest (ROI). In the recognition stage, the HSV color space is employed along with an extended twin support vector machines (TWSVMs) model to achieve the recognition of multi-type traffic lights including the arrow shapes. The dataset, collected in naturalistic driving experiments with an instrument vehicle, is utilized to train, verify, and evaluate the proposed method. The results suggest that compared with the previous YOLOv5s-based TLDR methods, the model size of the proposed lightweight TLDR method is reduced by 73.3%, and the computing power consumption of it is reduced by 79.21%. Meanwhile, the satisfied reasoning speed and recognition robustness are also achieved. The feasibility of the proposed method to be deployed on mobile platforms is verified with the Nvidia Jetson NANO platform.

Comprehensive, accurate, and timely cognition of navigation situation is a prerequisite to make safe, efficient, and scientific decisions for unmanned ships. Previous research on this topic has been more focused on situation perception, and lacks studies of mathematical representation, modeling, and prediction for multi-ship complex encounter situation from an integrated perspective. Based on this, vectorized model is constructed to express the situation for the target ship in open waters. Through analysis of bounded rational dynamic game between ships, the navigation situation of the target ship can be predicted. Historical data of Automatic Identification System (AIS) both from public data set and experiments is used to verify the prediction method. The results show that the research can provide theoretical basis for studies on multi-ship encounter situation, and supply a reference for navigation decisions of unmanned ships.

As one of the key technologies restricting the development of intelligent ships, autonomous collision avoidance has attracted the attention of many scholars all over the world. Existing research on collision-avoidance behavior focuses more on collision risk assessment and local path-planning methods for studies on the human-like sequential logic of the whole collision-avoidance process, as well as the decision-making process of various stages. Further in-depth thinking is needed urgently. Based on this, a construction method of a human-like sequential decision chain for the autonomous collision avoidance of unmanned ships is proposed through the construction of a collision-avoidance rule base and strategy set, efficient data access based on the Knowledge Graph concept, global collision risk assessment considering sequential decision process, and the construction of a complete collision-avoidance logic process to simulate the decision-making process of humans in complex multi-ship encounters in open waters. For multi-ship encounter scenarios, considering the sequential decision-making process of collision avoidance, a method was proposed to divide the collision risk of the target ship into direct collision risk and potential collision risk. The validity and reliability of the constructed sequential decision chain are verified by simulation experimental results. The results show that the method is effective for collision avoidance (especially multi-ship collision avoidance) in open waters and can provide a theoretical basis and technical support with good interpretability for the decision-making process of an unmanned ship’s autonomous collision avoidance.

为提升汽车主动安全功能,研究了1种基于高德导航数据的低成本、高精度驾驶倾向性辨识方法.基于高德软件开发工具构建动态驾驶数据采集应用程序,并融入个人智能终端以实现对行车数据的实时采集、处理与网络化存储.通过驾驶员生理、心理测试和实车实验获取不同驾驶倾向性驾驶员在导航行驶过程中由时间、速度和加速度推演的驾驶行为信息,采用主成分分析法(PCA)提取驾驶倾向性主要因子,并将驾驶倾向分为激进型、普通型和保守型这3类.构建基于果蝇优化算法(FOA)和广义回归神经网络(GRNN)的高精度驾驶倾向性辨识模型,利用特征变量集对模型进行训练和验证.验证结果表明:该模型总体准确率可达94.17％,对激进型、普通型和保守型的驾驶倾向性的辨识精确度分别为95.06％,92.5％,94.93％;进一步对比发现,该模型比单一的GRNN模型总体准确率提高5％～10％,与现有基于惯性传感器数据和离散小波变换结合自适应神经模糊推理系统的方法相比,该方法更具实用性且模型总体辨识准确率提升了2.17％.

Car-following behavior is the result of the interaction of various elements in the specific driver-vehicle-environment aggregation. Under the intelligent and connected condition, the information perception ability of vehicles has been significantly enhanced, and abundant information about the driver-vehicle-environment factors can be obtained and utilized to study car-following behavior. Therefore, it is necessary to comprehensively take into account the driver-vehicle-environment factors when modeling car-following behavior under intelligent and connected conditions. While there are a considerable number of achievements in research on car-following behavior, a car-following model with comprehensive consideration of driver-vehicle-environment factors is still absent. To address this gap, the literature with a focus on car-following behavior research with consideration of the driver, vehicle, or environment were reviewed, the contributions and limitations of the previous studies were analyzed, and the future exploration needs and prospects were discussed in this paper. The results can help understand car-following behavior and the traffic flow characteristics affected by various factors and provide a reference for the development of traffic flow theory towards smart transportation systems and intelligent and connected driving.

Aiming at the problems of low detection accuracy and poor recognition effect of small-scale targets in traditional vehicle and pedestrian detection methods, a vehicle and pedestrian detection method based on improved YOLOv4-Tiny is proposed. On the basis of YOLOv4-Tiny, the 8-fold down sampling feature layer was added for feature fusion, the PANet structure was used to perform bidirectional fusion for the deep and shallow features from the output feature layer of backbone network, and the detection head for small targets was added. The results show that the mean average precision of the improved method has reached 85.93%, and the detection performance is similar to that of YOLOv4. Compared with the YOLOv4-Tiny, the mean average precision of the improved method is increased by 24.45%, and the detection speed reaches 67.83FPS, which means that the detection effect is significantly improved and can meet the real-time requirements.

Emotion is an implicit psychological characteristic that changes over time. When it accumulates to a certain extent, it will be accompanied by certain external manifestations. Drivers with different traits have different emotional performance, which leads to different effects from different driver traits on the driver’s emotional activation efficacy. In this study, we thoroughly explore the effects of different genders, age, driving competence, driving anger tendency, driving safety attitude and stress state on driver’s emotional activation efficacy. This paper selects 74 young and middle-aged drivers with an age distribution between 20 and 41 years old. The eight most typical driving emotions (anger, surprise, fear, anxiety, helplessness, contempt, ease and pleasure) were screened through questionnaires. An experimental framework for the emotional stimulation and measurement of eight driving emotions was designed based on multiple emotional stimulation methods and PAD emotional model. The effect of emotional activation on drivers of different genders, age, driving competence, driving anger tendency, driving safety attitude and stress state was explored in depth. The results show that gender, age, driving safety attitude, driving anger tendency, stress state, etc., all have different degrees of influence upon the activation efficacy of emotion. The research results reveal the rules for the generation of different driving emotions to a certain extent and provide a theoretical basis for further exploring the cognitive and behavioral characteristics of drivers with different emotions.

Aggressive driving behavior (ADB) is one of the main causes of traffic accidents. The accurate recognition of ADB is the premise to timely and effectively conduct warning or intervention to the driver. There are some disadvantages, such as high miss rate and low accuracy, in the previous data-driven recognition methods of ADB, which are caused by the problems such as the improper processing of the dataset with imbalanced class distribution and one single classifier utilized. Aiming to deal with these disadvantages, an ensemble learning-based recognition method of ADB is proposed in this paper. First, the majority class in the dataset is grouped employing the self-organizing map (SOM) and then are combined with the minority class to construct multiple class balance datasets. Second, three deep learning methods, including convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU), are employed to build the base classifiers for the class balance datasets. Finally, the ensemble classifiers are combined by the base classifiers according to 10 different rules, and then trained and verified using a multi-source naturalistic driving dataset acquired by the integrated experiment vehicle. The results suggest that in terms of the recognition of ADB, the ensemble learning method proposed in this research achieves better performance in accuracy, recall, and F1-score than the aforementioned typical deep learning methods. Among the ensemble classifiers, the one based on the LSTM and the Product Rule has the optimal performance, and the other one based on the LSTM and the Sum Rule has the suboptimal performance.

It is of great practical and theoretical significance to identify driver fatigue state in real time and accurately and provide active safety warning in time. In this paper, a non-invasive and low-cost method of fatigue driving state identification based on genetic algorithm optimization of generalized regression neural network model is proposed. The specific work is as follows: (1) design simulated driving experiment and real driving experiment, determine the fatigue state of drivers according to the binary Karolinska Sleepiness Scale (KSS), and establish the fatigue driving sample database. (2) Improved Multi-Task Cascaded Convolutional Networks (MTCNN) and applied to face detection. Dlib library was used to extract the coordinate values of face feature points, collect the characteristic parameters of driver's eyes and mouth, and calculate the Euler Angle parameters of head posture. A fatigue identification model was constructed by using multiple characteristic parameters. (3) Genetic Algorithm (GA) was used to find the optimal smooth factor of Generalized Regression Neural Network (GRNN) and construct GA-GRNN fatigue driving identification model. Compared with K-Nearest Neighbor (KNN), Random Forest (RF), and GRNN fatigue driving identification algorithms. GA-GRNN has the best generalization ability and high stability, with an accuracy of 93.3%. This study provides theoretical and technical support for the application of driver fatigue identification.

Driving propensity is the driver's attitude towards the actual traffic situation and the corresponding decision-making or behavior during the driving process. It is of great significance to improve the accuracy of safety early warning and reduce traffic accidents. In this paper, a real-time identification system of driving propensity based on AutoNavi navigation data is proposed. The main work includes: (1) A dynamic data acquisition method of AutoNavi navigation is proposed to obtain the time, speed and acceleration of the driver during the navigation process. (2) The dynamic data collection method of AutoNavi navigation is analyzed and verified through the dynamic data obtained in the real vehicle experiment. The principal component analysis method is used to process the experimental data to extract the driving propensity characteristics variables. (3) The fruit fly optimization algorithm combined with GRNN (generalized neural network) and the feature variable set are used to build a FOA-GRNN-based model. The results show that the overall accuracy of the model can reach 94.17%. (4) A driving propensity identification system is constructed. The system has been verified through real vehicle test experiments. This paper provides a novel and convenient method for building personalized intelligent driver assistance systems in practical applications.

The visual attention system is the gateway to the human information processing system, and emotion is an important part of the human perceptual system. In this paper, the driver's visual attention characteristics and the influences of typical driving emotions on those were explored through analyzing driver's fixation time and identification accuracy to different visual cognitive tasks during driving. The results showed that: the increasing complexity of the cognitive object led to the improvement of visual identification speed. The memory and recall process increased drivers' fixation time to cognitive objects, and the recall accuracy decreased with the increase in time interval. The increase in the number of cognitive objects resulted in the driver improving the visual identification speed for the cognitive object at the end of the sequence consciously. The results also showed that: the visual cognitive efficiency was improved in the emotional states of anger and contempt, and was decreased in the emotional states of surprise, fear, anxiety, helplessness and pleasure, and the emotional state of relief had no significant effect on the visual cognitive efficiency. The findings reveal the driver's visual information processing mechanism to a certain extent, which are of great significance to understand the inner micro-psychology of driver's cognition.

The application of vehicle-to-everything (V2X) technology has resulted in the traffic environment being different from how it was in the past. In the V2X environment, the information perception ability of the driver–vehicle unit is greatly enhanced. With V2X technology, the driver–vehicle unit can obtain a massive amount of traffic information and is able to form a connection and interaction relationship between multiple vehicles and themselves. In the traditional car-following models, only the dual-vehicle interaction relationship between the object vehicle and its preceding vehicle was considered, making these models unable to be employed to describe the car-following behavior in the V2X environment. As one of the core components of traffic flow theory, research on car-following behavior needs to be further developed. First, the development process of the traditional car-following models is briefly reviewed. Second, previous research on the impacts of V2X technology, car-following models in the V2X environment, and the applications of these models, such as the calibration of the model parameters, the analysis of traffic flow characteristics, and the methods that are used to estimate a vehicle’s energy consumption and emissions, are comprehensively reviewed. Finally, the achievements and shortcomings of these studies along with trends that require further exploration are discussed. The results that were determined here can provide a reference for the further development of traffic flow theory, personalized advanced driving assistance systems, and anthropopathic autonomous-driving vehicles.

The type of vehicles in realistic traffic systems are not homogeneous. Impacts of the preceding vehicle’s type on the car-following behavior, fuel consumption and exhaust emissions are still unclear. This paper presents a study on the impacts of two types of preceding vehicles, heavy vehicles and new energy vehicles, on car-following behavior, traffic flow characteristics, fuel consumption and exhaust emissions. Firstly, an extended car-following model was proposed by incorporating the influence of the preceding vehicle’s type. Secondly, impacts of the preceding vehicle’s type on platoon stability were analyzed by applying linear stability theory. Finally, numerical simulations were carried out to analyze impacts of the preceding vehicles’ type on the characteristics of the car-following platoon, traffic flow operation, and vehicle’s fuel consumption and exhaust emissions. The results reveal that, compared with the normal preceding vehicle, there are negative impacts of the heavy and new-energy preceding vehicles on the platoon stability, traffic flow operation, and vehicle’s fuel consumption and exhaust emissions, and these impacts are related to the corresponding sensitivity parameters and the penetration percentages of the two types of preceding vehicle. The research results of this paper can provide a reference for understanding car-following behavior and traffic-flow characteristics affected by the type of preceding vehicles in the V2X environment.

Accurate identification of abnormal driving behavior is very important to improve driver safety. Aiming at the problem that threshold or traditional machine learning methods are mostly used in existing studies, it is difficult to accurately identify abnormal driving behavior of vehicles, a method of abnormal driving behavior recognition based on smartphone sensor data and convolutional neural network (CNN) combined with long and short-term memory (LSTM) was proposed. Smartphone sensors are used to collect vehicle driving data, and data sets of various driving behaviors are constructed by preprocessing the data. A recognition model based on a convolutional neural network combined with a long short-term memory network was constructed to extract depth features from data sets and recognize abnormal driving behaviors. The test results show that the accuracy of the model based on CNN-LSTM can reach 95.22%, and the performance indexes can reach more than 94%. Compared with the recognition model constructed only by CNN or LSTM, this model has higher recognition accuracy.

The car-following models are the research basis of traffic flow theory and microscopic traffic simulation.Among the previous work, the theory-driven models are dominant, while the data-driven ones are relatively rare.In recent years, the related technologies of Intelligent Transportation System (ITS) represented by the Vehicles to Everything (V2X) technology have been developing rapidly.Utilizing the related technologies of ITS, the large-scale vehicle microscopic trajectory data with high quality can be acquired, which provides the research foundation for modeling the car-following behavior based on the data-driven methods.According to this point, a data-driven car-following model based on the Random Forest (RF) method was constructed in this work, and the Next Generation Simulation (NGSIM) dataset was used to calibrate and train the constructed model.The Artificial Neural Network (ANN) model, GM model, and Full Velocity Difference (FVD) model are employed to comparatively verify the proposed model.The research results suggest that the model proposed in this work can accurately describe the carfollowing behavior with better performance under multiple performance indicators.