To enable the inference of high-precision deep neural networks (DNNs) on resource-constrained devices, DNN offloading has been widely explored in recent years. Some works have also integrated the chain-topology DNN (CDNN) offloading with pipeline processing to further reduce inference delay when processing streaming tasks. To improve the accuracy of the inference results, the topology of DNN tends to evolve from chain topology to directed acyclic graph (DAG) topology. However, most of the existing works do not study partitioning and offloading DAG-topology DNNs (DDNNs) for streaming tasks. Moreover, when partitioning computationally expensive DNN models, multipartitioning probably outperforms the bi-partitioning method, and most of the works do not study multipartitioning DAG-topology DNNs. In this article, we propose a more general multipartitioning and offloading method for large-scale DDNNs to process streaming tasks, which can adaptively partition DDNNs into multiple parts considering the computing power and bandwidth of all available computing units. Specifically, we first present a transforming method based on topological sorting that can losslessly transform DAG-topology DNNs into CDNNs. Then, based on greedy and dichotomy ideas, a multipartitioning algorithm is designed to partition and offload CDNNs. In this way, we can solve DDNNs’ multipartitioning problem based on the proposed transforming and partitioning algorithms. Experiments https://github.com/sreasearcher/PDD-Code show that the method proposed in this article significantly outperforms bi-partitioning and nonpartitioning methods when offloading computationally expensive DNN models.

A concrete dam is an important water-retaining hydraulic structure that stops or restricts the flow of water or underground streams. It can be regarded as a constantly changing complex system. The deformation of a concrete dam can reflect its operation behaviors most directly among all the effect quantities. However, due to the change of the external environment, the failure of monitoring instruments, and the existence of human errors, the obtained deformation monitoring data usually miss pieces, and sometimes the missing pieces are so critical that the remaining data fail to fully reflect the actual deformation patterns. In this paper, the composition, characteristics, and contamination of the concrete dam deformation monitoring information are analyzed. From the single-value missing data completion method based on the nonlocal average method, a multi-value missing data completion method using BP (back propagation) mapping of spatial adjacent points is proposed to improve the accuracy of analysis and pattern prediction of concrete dam deformation behaviors. A case study is performed to validate the proposed method.

A ship collision accident is one of the most dangerous and common types of maritime accidents. Traditional probabilistic risk assessment (PRA) of ship collision accidents is a methodology that can be adopted to ensure maritime safety. Nevertheless, a need for better approaches to model human behavior, such as risk identification, communication, and decision-making, has been identified. Such advanced PRA methods require a more explicit way of taking human factors into consideration than the traditional risk assessment methods. Hybrid causal logic (HCL) is an advanced PRA method due to its unique three-level framework that includes event sequence diagrams, fault trees, and Bayesian networks, which makes it suitable for modeling human behavior that is important to ship collision accidents. This paper discusses the applicability of the HCL methodology for the ship collision accident. Firstly, the event sequences of typical ship collision accidents are summarized based on the study of 50 accident investigation reports. Then, fault trees for mechanical failure events and the Bayesian networks for human error events are constructed to analyze the events in a structured way at a more detailed level. Finally, the three main end-state types of ship collision avoidance scenario have been quantified. The result of the probability of a ship collision accident is verified by estimating the annual frequency of collision accidents in the Singapore Strait. Comparing with the historical data, the estimation results are quite near to the real case. By taking advantage of the HCL methodology, the modeling of ship collision scenarios can be carried out at a deep logical level. At the same time, it is possible to combine a detailed analysis of various primary events with a comprehensive analysis at the system level.

In order to monitor traffic in congested waters, permanent video stations are now commonly used on interior riverbank bases. It is frequently challenging to identify ships properly and effectively in such images because of the intricate backdrop scenery and overlap between ships brought on by the fixed camera location. This work proposes Ship R-CNN(SR-CNN), a Faster R-CNN-based ship target identification algorithm with improved feature fusion and non-maximum suppression (NMS). The SR-CNN approach can produce more accurate target prediction frames for prediction frames with distance intersection over union (DIOU) larger than a specific threshold in the same class weighted by confidence scores, which can enhance the model’s detection ability in ship-dense conditions. The SR-CNN approach in NMS replaces the intersection over union (IOU) filtering criterion, which solely takes into account the overlap of prediction frames, while DIOU, also takes into account the centroid distance. The screening procedure in NMS, which is based on a greedy method, is then improved by the SR-CNN technique by including a confidence decay function. In order to generate more precise target prediction frames and enhance the model’s detection performance in ship-dense scenarios, the proposed SR-CNN technique weights prediction frames in the same class with DIOU greater than a predetermined threshold by the confidence score. Additionally, the SR-CNN methodology uses two feature weighting methods based on the channel domain attention mechanism and regularized weights to provide a more appropriate feature fusion for the issue of a difficult ship from background differentiation in busy waters. By gathering images of ship monitoring, a ship dataset is created to conduct comparative testing. The experimental results demonstrate that, when compared to the three traditional two-stage target detection algorithms Faster R-CNN, Cascade R-CNN, and Libra R-CNN, this paper’s algorithm Ship R-CNN can effectively identify ship targets in the complex background of far-shore scenes where the distinction between the complex background and the ship targets is low. The suggested approach can enhance detection and decrease misses for small ship targets where it is challenging to distinguish between ship targets and complex background objects in a far-shore setting.

Human error is a crucial factor leading to maritime traffic accidents. The effect of human–computer interaction (HCI) also plays a leading role in human error. The objective of this study is to propose a method of interaction strategies based on a cognitive-processing model in crews’ daily navigation tasks. A knowledge-based ship HCI framework architecture is established. It provides an extensible framework for the HCI process in the maritime domain. By focusing on the cognitive process of a crew in the context of accident and risk handling during ship navigation, based on the information, decision, and action in crew context (IDAC) model, in combination with the maritime accident dynamics simulation (MADS) system, the MADS-IDAC system was developed and enhanced by the HCI structure and function design of the dynamic risk analysis platform for maritime management. The results indicate that MADS enhanced by HCI can effectively generate a strategy set of various outcomes in preset scenarios. Moreover, it provides a new method and thought for avoiding human error in crew interaction and to lower the risk of ship collision as well as effectively improving the reliability of HCI.

It is expected that the prototypes of unmanned merchant ships will be deployed in the next few years. However, there is no specific research on whether the introduction of unmanned ships will reduce the risk of ship collision accidents in which communication between vessels is critical. This work constitutes an attempt to bridge the gap identified above by applying the Hybrid Causal Logic (HCL) methodology to model general-level collision scenarios of unmanned ships. The HCL methodology has been selected for its proven applicability to risk assessments, even when empirical data may be insufficient. Collision scenarios involving unmanned ships have been created in which manned ships of the conventional collision scenario HCL model are replaced with unmanned ships. Then, collision scenarios capturing the interactions between a manned ship and an unmanned ship were modeled. By comparing the qualitative and quantitative results of the different scenarios, we can see that the introduction of unmanned ships may effectively reduce the occurrence of ship collision accidents.