In light of the limitations of existing object detection models in extracting complex features of oil spill areas and identifying irregularly shaped oil spill regions, an improved Mask R-CNN model is proposed to better adapt to the task of oil spill detection on the sea surface. Firstly, a deformable convolution module is introduced into the feature extraction network of the model to enhance its perception of irregularly shaped oil spill areas. Secondly, an improved attention mechanism is incorporated into the model, which not only strengthens the model's ability to capture key features of oil spill areas but also adds relatively few parameters. Lastly, the Complete Intersection over Union Loss (CIoU Loss) function is utilized as the bounding box regression loss function, thereby improving the model's performance in the bounding box regression task. Experiments conducted on a publicly available Synthetic Aperture Radar (SAR) sea surface oil spill dataset demonstrate that the detection accuracy of the improved model is 66.14%, representing an increase of 5.33% compared to the original Mask R-CNN model. Moreover, on the same dataset, when compared with object detection models Yolov9, Yolov10, Faster R-CNN, and Cascade R-CNN, the accuracy is respectively enhanced by 31.79%, 19.01%, 30.47%, and 21.02%. In comparison with instance segmentation models Yolact and Yolov5-seg, the accuracy is respectively elevated by 22.94% and 29.5%.

This paper proposes an adaptive correction and feature enhancement image preprocessing algorithm to address the problem of feature point matching failure caused by uneven lighting and blurry perceived image texture during autonomous navigation of unmanned ships. Firstly, Unsharp Masking was used to enhance image details. Then, multi-scale Gaussian convolution was used to extract lighting components in the brightness channel, and two-dimensional gamma correction was used to homogenize brightness. Gaussian filtering was applied to the tone channel to suppress low-pass noise. Finally, histogram homogenization algorithm was used to improve image contrast. Compared with the MSR algorithm, MSRCR algorithm, and SSR algorithm, the average gradient of this algorithm has increased by up to 50.55%, 151.21%, and 43.68% respectively, and the feature matching accuracy has increased by up to 86.81%, 176.08%, and 61.96% respectively. This research provides preprocessing techniques for visual perception images in autonomous navigation of unmanned ships, improving image quality and having certain application value.

 In the online order picking system of large supermarkets, considering the workload balance among pickers, the optimization of order batching and picking route was studied. For the scenario of " sort-after-pick" batch picking in large supermarkets with a limited number of pickers during peak periods, a dual-objective optimization model was constructed with two objectives of minimizing total completion time and minimizing the range in completion times. According to the problem feature and the dual-objective solution method, an improved NSGA-II algorithm was designed by combining the K-means clustering algorithm and the nearest neighbor strategy of the greedy algorithm. Based on the practical operation of large supermarkets such as Walmart, a picking layout and case parameters were set. The correctness of the model and the effectiveness of the algorithm were verified through examples of different scales. Numerical experiments show that range as workload balance criteria in a dual-objective model not only achieve workload balance but also has fewer negative impacts on picking efficiency. A further sensitivity analysis indicates that a population size of 50 and an iteration number of 100 are conducive to obtaining quality solutions. Comparative experiments conducted on datasets of different sizes reveals that the "sort-after-pick" method can reduce the average completion time of orders by 44.37% in comparison to the single order picking strategy. The conclusion indicates that the dual-objective model and algorithm can achieve a balance between workload balance and picking efficiency, improving picking efficiency while satisfying workload balance requirement from pickers’ perspective.

Polyvinyl alcohol hydrogel flexible surface is used as drag reducing material, and sodium alginate and hydroxypropyl methyl cellulose are added to enhance its elasticity and mechanical strength. Test the drag reduction effect of different spacing flow direction groove flexible surfaces and vertical flow direction groove flexible surfaces using the rotating disk drag measurement method, and analyze their impact mechanism on the disk drag reduction effect. The results show that the hydrogel flexible surface without grooves has a good drag reduction effect in laminar flow, but it will lose the drag reduction effect in turbulent flow and play the role of drag increase. The two kinds of hydrogel flexible surfaces with sparse grooves have higher drag reduction rate in laminar flow, and can also maintain drag reduction effect in high Reynolds number turbulent state, which don’t have the effect of increasing resistance.

Efficient tank washing is a key process to ensure the efficiency and quality of chemical transportation. In order to study the factors affecting the washing effect of chemical tankers, this paper selected palm oil as a typical non-volatile chemical cargo, took a stainless steel lined bulkhead chemical tanker as a research object, built a physical experiment system and developed a reliable numerical simulation model. During the experiment, it was found that the effect of tank washing in the early stage is mainly affected by the water jet, and in the middle and later stage it is mainly due to the dissolution of tank washing water. The research results show that the main factors affecting the tank washing effect are the length and diameter of the nozzle outlet, the washing time, the temperature of the tank washing water and the dynamic pressure. The quantitative evaluation model of palm oil tank washing effect is obtained by analyzing the results. This paper has theoretical reference value for related research and good guidance significance for the practice of chemical tanker washing.

In order to accurately reflect the nonlinear, stochastic, and non-stationary characteristics of ship pitch motion in real-time, an adaptive prediction model is proposed based on sliding data window and Lipschitz quotients method. Firstly, the sliding data window is employed as a local observer to segment the ship's pitch motion status data in real-time, and the Lipschitz quotients method is used to adaptively determine the order of the subsystems represented within the current sliding data window. Online small batch training samples are provided for the feed-forward neural network model by using sliding data window and Lipschitz quotients method, which can overcome the impact of single sample and big batch data samples on the performance of the neural network model. Then, to address the problem that feed-forward neural networks based on deterministic learning algorithms are prone to fall into local optimums, a feed-forward neural network model based on the improved butterfly optimization algorithm trainer is proposed to improve the prediction accuracy of the ship's pitch motion status. In the improved butterfly optimization algorithm, a mutation operator guided by the balancing factor and an information reorganization strategy with an optimal individual guidance mechanism are employed to enhance the algorithm's ability to avoid falling into a local optimum. Finally, the effectiveness and feasibility of the improved butterfly optimization algorithm and the adaptive prediction model are verified by using the benchmark test functions and the measured pitch motion status data from M.V. “YuKun”, respectively. The experimental results show that the improved butterfly optimization algorithm in respect of convergence speed and accuracy outranks the butterfly optimization algorithm, particle swarm optimization algorithm, and moth-flame optimization algorithm; The proposed adaptive prediction model has stronger generalization ability and higher prediction accuracy, and the average running time of each step is within 0.2s, which is less than the system sampling time of 1s. The proposed adaptive prediction model not only meets real-time requirements but also improves the accuracy of ship's pitch motion status prediction, which can provide a potential solution for online modeling of complex systems.

Ship trajectory prediction is essential for intelligent ships to understand complex encounter scenarios and make wise decisions. However, due to inherent uncertainty and complex interactions between different ships, predicting future trajectories is a very challenging problem. Therefore, a ship trajectory prediction model based on Multi-relational weighted graph Transformer (MG-Transformer) is proposed. First, the motion patterns of ships with similar trajectories are extracted from AIS data to capture different movement features. On this basis, the historical motion patterns of different ships are learned to improve the prediction accuracy and efficiency of the model. Secondly, the multi-relational weighted graphs is constructed to illustrate the complex spatial relationship between multiple ships. The interaction with surrounding ships is learned through Transformer to refine the trajectory and predict a reasonable trajectory. The AIS data of Ningbo-Zhoushan Port is used for experimental verification. The results show that when predicting trajectories of different time steps, compared with LSTM, BiLSTM, Seq2seq, and Social-SGCNN, the MG-Transformer model has a significant decrease in the average displacement error and final displacement error indicators. The average reduction of each indicator is 27.54%. The accuracy of the proposed ship trajectory prediction model has been significantly improved, which is crucial for maritime traffic safety and efficiency. 

To enhance the stress-adaptive characteristics of the FeMnSi shape memory alloy's γ↔ε martensitic transformation, improve its fatigue strength, wear resistance, residual stress release, stress concentration reduction, and microcrack inhibition capabilities, this paper studies the process and performance characteristics of FeMnSiCrNi shape memory alloy coatings prepared by laser alloying on the surface of 316 stainless steel. The study uses laser alloying technology to prepare FeMnSiCrNi shape memory alloy coatings on the surface of 316 stainless steel. The shape and size of the coating molten pool are simulated using the finite element analysis software ANSYS. After optimizing the laser alloying process parameters, the best process is selected as a laser power of 2000W, a scanning speed of 400 mm/s, a defocusing distance of -30 mm, and an overlap rate of 50%. Subsequently, the microstructure, residual stress distribution, mechanical properties, and wear resistance of the coating are systematically analyzed using a scanning electron microscope (SEM), X-ray diffractometer (XRD), X-ray stress analyzer, microhardness tester, and friction tester. The observation results show that the coating structure is dense, the surface is smooth, and it forms a good metallurgical bond with the 316 stainless steel substrate. It is mainly composed of γ austenite phase and a small amount of ε martensite phase. The residual stress generated during the laser alloying process induces the γ→ε martensitic transformation. After the coating cools, the transverse residual stress in the middle area is compressive stress, and it gradually changes to tensile stress on both sides, showing a "compressive stress→tensile stress→compressive stress" distribution along the laser scanning direction. The hardness of the FeMnSiCrNi shape memory alloy coating is significantly higher than that of the 316 stainless steel substrate, and the friction coefficient is lower. Under dry friction conditions, at loads of 10N, 15N, and 20N, the friction coefficients of the Fe17Mn5Si10Cr5Ni coating are 0.46, 0.57, and 0.97, respectively, while those of the stainless steel substrate are 0.57, 0.98, and 1.33, respectively. Under dry friction for 10 minutes, the wear amounts of the Fe17Mn5Si10Cr5Ni coating are 0.17g (10N load), 0.29g (15N load), and 0.50g (20N load), significantly lower than those of the 316 stainless steel substrate, which are 0.42g (10N load), 0.81g (15N load), and 1.12g (20N load), respectively. The wear mechanism of the FeMnSi shape memory alloy coating is abrasive wear, while the 316 stainless steel substrate mainly shows adhesive wear. The test results show that the Fe17Mn5Si10Cr5Ni shape memory alloy coating prepared by laser alloying technology exhibits excellent mechanical properties and wear resistance, and verifies the important role of the γ→ε martensitic transformation in optimizing the coating performance. This coating not only significantly improves the hardness and wear resistance of 316 stainless steel, but also optimizes the friction coefficient and residual stress distribution, providing a new theoretical basis and practical solution for the design of high-performance FeMnSi shape memory alloy materials and metal surface modification. 

The hydrodynamics of four-column structures in waves is studied by using AQWA software. The relationships of the first order wave force and the second order wave force with wave frequency are analyzed, and the wave elevations at the characteristic points around the cylinder and the wave surface distribution around the cylinder at different frequencies are given. The results of research shows the following conclusions: When kr < 0.5 in relatively low frequency region, the peak values of the first order surge force of each cylinder do not show much differences; however, the first-order longitudinal wave forces of the cylinder show obvious oscillating characteristics in relatively middle and high frequency region when  kr ＞ 0.5, the maximum peak value is reached near kr = 1.68, but the difference of the first order heave wave force on each column is not obvious. The second order wave force of the four-column structure is directed from inside of the structure to the outside of the structure. With the increase of wavelength, the transmitted ability of the wave passing through the four-column body becomes stronger. However, when kr is around 1.68, the wave interference resonance occurs in the internal space of the four-column body, resulting in a significant rise of the wave around the column. This study reveals the law of hydrodynamic interference and resonance phenomenon between multiple cylinders in waves, which can provide theoretical reference for the subsequent design of pile foundation of offshore wind turbine platform.