The quality of Automatic Identification System (AIS) data is of fundamental importance for maritime situational awareness and navigation risk assessment. To improve operational efficiency, a deep learning method based on Bi-directional Long Short-Term Memory Recurrent Neural Networks (BLSTM-RNNs) is proposed and applied in AIS trajectory data restoration. Case studies have been conducted in two distinct reaches of the Yangtze River and the capability of the proposed method has been evaluated. Comparisons have been made between the BLSTM-RNNs-based method and the linear method and classic Artificial Neural Networks. Satisfactory results have been obtained by all methods in straight waterways while the BLSTM-RNNs-based method is superior in meandering waterways. Owing to the bi-directional prediction nature of the proposed method, ship trajectory restoration is favourable for complicated geometry and multiple missing points cases. The residual error of the proposed model is computed through Euclidean distance which decreases to an order of 10 m. It is considered that the present study could provide an alternative method for improving AIS data quality, thus ensuring its completeness and reliability.

A hybrid-clustering model is presented for the probabilistic characterization of ship traffic and anomaly detection. A hybrid clustering model was proposed to increase the efficiency of trajectory clustering in the port area and analyze the maritime traffic patterns in port. The model identified dissimilarities between trajectories based on characteristics, using K-Means and the density-based spatial clustering of applications with noise algorithm (DBSCAN). Firstly, the ship’s trajectory characteristics are constructed based on real ship trajectories considering static characteristics and dynamic characteristics of ship trajectories to calculate the characteristic dissimilarity between trajectories. Simultaneously, the spatial dissimilarity could be quantified using the Hausdorff algorithm. Then, the ship trajectory is clustered initially based on the departure and destination characteristics using K-Means algorithms to obtain various sub-trajectories. However, there are still different types of trajectories in each sub-trajectory. Thus, the DBSCAN algorithm is adopted to cluster the sub-trajectory based on the analysis of the different trajectory characteristics. Finally, the proposed model is applied to the characterization of the Zhanjiang Port, and the results show that the hybrid-clustering method can effectively cluster ship trajectory and present probabilistic characterization of ship traffic and anomaly detection. This lays a solid theoretical foundation for the supervision and risk control of intelligent ships.

COVID-19 first appeared in Wuhan city of Hubei Province in China in December 2019. It has a substantial impact on human life all around the world, especially for citizens. The threat of COVID-19 has resulted in people shopping online to get fresh food and reduce outdoor trips. Collecting data from adult internet users in Wuhan, China in 2020, this study aims to explore the influence of COVID-19 on fresh food shopping behavior. In addition, a comparison and ordered logit model are constructed to demonstrate the changes and effects of COVID-19. The results suggest that more citizens in Wuhan city will buy fresh food online and the cost and frequency are also increased. The experience of online shopping for fresh food during the lock-down days has promoted more online shopping. The factors, such as frequency of online shopping before the COVID-19 outbreak, frequency of online shopping during the COVID-19 pandemic, and age, have a negative effect on the proportion of online shopping after the lock-down days, while the proportion of online shopping before the COVID-19 outbreak, the proportion of online shopping during the COVID-19 pandemic, and travel time of in-store shopping before the COVID-19 outbreak have a positive effect. The results provide insights for managers, city planners, and policymakers.

The ship safety domain plays a significant role in collision risk assessment. However, few studies take the practical considerations of implementing this method in the vicinity of bridge-waters into account. Therefore, historical automatic identification system data is utilised to construct and analyse ship domains considering ship–ship and ship–bridge collisions. A method for determining the closest boundary is proposed, and the boundary of the ship domain is fitted by the least squares method. The ship domains near bridge-waters are constructed as ellipse models, the characteristics of which are discussed. Novel fuzzy quaternion ship domain models are established respectively for inland ships and bridge piers, which would assist in the construction of a risk quantification model and the calculation of a grid ship collision index. A case study is carried out on the multi-bridge waterway of the Yangtze River in Wuhan, China. The results show that the size of the ship domain is highly correlated with the ship's speed and length, and analysis of collision risk can reflect the real situation near bridge-waters, which is helpful to demonstrate the application of the ship domain in quantifying the collision risk and to characterise the collision risk distribution near bridge-waters.

Ship detection and tracking has been recognized as a challenging task in the maritime administration. The method of LSDT (Light Spot Detection and Tracking) has been applied to achieve ship detection and tracking based on the nighttime surveillance video. Firstly, the light spots in the video images are detected through LOG and invalid spots are filtered by the gray threshold. Multiple targets are subsequently tracked by Kalman filtering and light spots are marked to determine properties in order to add and delete spots. A case study has been performed in the middle reach (Wuhan) of the Yangtze River. Nighttime surveillance video has been obtained and further utilized for method verification. The results show that the ship can be detected and tracked by LSDT effectively. The present study provides an alternative for ship detection at night, thus improving the maritime surveillance efficiency.