To solve the problem of the large number of parameters and computation of ship target detection algorithm, as well as the difficulties of ship detection caused by the influence of the nearshore complex backgrounds and the mutual occlusion of ships in inland river environments, this paper makes improvements based on YOLOv7-tiny and proposes a lightweight algorithm MED-YOLO for ship target detection. Firstly, the MobileNetV3 network is used as the backbone feature extraction network, which greatly reduces the calculation cost of the model. Secondly, EMA attention module was introduced into the neck network, and EMA-ELAN module was constructed to enhance the multi-dimensional perception and multi-scale feature extraction capability of the network. Then, Dyhead, which combines scale perception, spatial perception, and task perception, is selected as the detection head of the improved model to obtain stronger feature expression ability. Finally, WIoU with dynamic non-monotonic focusing mechanism is used as the bounding box loss function to improve the model's ability to cope with ship occlusion and improve the detection performance. The experimental results show that compared with YOLOv7-tiny, MED-YOLO has 39.8% fewer parameters and 55.0% less computation, and its precision and mAP@0.5 have increased by 1.4% and 1.0% respectively, reaching 98.3% and 98.9%, which not only achieves lightweight, but also has better detection performance. It meets the deployment requirements in the environment with limited computing resources, and has certain practical engineering significance.

According to the change law reflected in the Arctic shipping data, the characteristics of Arctic shipping are systematically arranged: the navigation period is summer and autumn, the shipping volume is constrained by external conditions, the shipping structure is single and external transportation is the mainstay. On this basis, it is proposed that the transportation time, transportation complexity and transportation range are more suitable for the pendulum transportation mode of Arctic shipping, and the pendulum transportation mode design of the northeast Arctic route is carried out from the aspects of fulcrum port, ship type and modular transportation, and the annual shipping revenue of the pendulum transportation mode and the traditional transportation mode and mixed transportation mode are comparatively analyzed. The results show that compared with the other two transport modes, the pendulum transportation mode of the Northeast Arctic route can complete one more voyage per year, and the annual shipping revenue is 9% and 11.3%, respectively. Arctic shipping can be developed according to this mode of transport to further develop the potential of Arctic shipping.

In view of the complexity of the ice situation and geographical environment of the Arctic shipping route, there are risks such as ship ice collision and grounding in polar navigation. In order to ensure the safety and economy of ship polar navigation, a route planning method that comprehensively considered multiple risk factors such as water depth, sea ice concentration and sea ice thickness was proposed. Firstly, for the global route planning, the polar operational limit assessment risk indexing system (POLARIS) was adopted to calculate the risk index outcome (RIO) of the Arctic safe water depth region, thereby reclassifying the risk grade and assigning it to risk cost. The lowest risk cost route in the safe water depth region was generated by geographic information system (GIS) lowest cost path analysis method and was recommended as the global route. Secondly, the grid proportion of each risk level was extracted to evaluate the feasibility of the route. Finally, for the sea ice area along the global route that was larger than the maximum ice thickness of the International Association of Classification Societies (IACS) polar class ship operating restrictions comparison table, the local route planning was carried out by combining with the improved A* and the dynamic window approach (DWA), thereby considering the risk of ship-ice collision under complex and variable ice conditions. The comparison of global and local routes with different risk factors shows that the global routes planned in this paper have the lowest risk with grounding avoidance, in addition that the ship can make collision-avoidance decision against the new ice blocking the channel when sailing along the local planned route.

Due to inherent scattering and absorption, underwater images inevitably suffer from multiple degradations arising from blurring, low contrast and color distortion, thereby seriously deteriorating visual perception. In this paper, a deep learning-based underwater image restoration and enhancement framework (UIRENet) was proposed by virtue of depth estimation and gradient descent strategy. With the aid of convolutional and nonlinear activation function modules, a deep perception network was constructed to achieve scene depth perception maps for different degradation regions, thereby overcoming the dependence of scene depth degradation. A gradient optimization strategy was further proposed to optimize the parameters of convolutional networks and improve the performance of deep network enhancement. Combined with perceptual, edge and underwater color constancy losses, a comprehensive loss function for underwater image enhancement networks was rationally formed.  Comprehensive experiments on the UIEB-90, UIEB-M and EUVP datasets show that the UIRENet framework significantly outperforms  typical underwater image enhancement methods in terms of reducing underwater image blurriness and improving visual effects. In particular, comparing to CLAHE, ICM, GC, IBLA, DCP, ULAP, FUnIE-GAN, UGAN and Uformer, the objective evaluation metric UIQM can be promoted by 0.3700, 0.6446, 0.5919, 1.3081,1.3032, 1.1672, 0.0593, 0.1329 and 0.0934, respectively.

As the necessary core equipment of ocean-going merchant ships, ship navpilot can reduce the burden of ship pilots, improve navigation safety and economy, and is the hot direction of intelligent ship development. In this paper, focuses on the robust adaptive track keeping algorithm as the core technology, considering the practical control requirements of navigation engineering and using embedded Linux technology, a set of navpilot prototype with navigation control, monitoring and alarm, system information display functions, hardware interface in line with the requirements of international standards is developed. Finally, taking "Yu-Kun" as the experimental object, the effectiveness of the ship navpilot was tested according to the Maritime navigation and radiocommunication equipment and systems-Track control systems-Operational and performance requirements，methods of testing and required test results (GB/T 37417-2019) national standard test scene under random Marine environment interference conditions. The results show that the ship navpilot can realize the stable tracking control of the course spanning 0° longitude and latitude, and the course error is less than 2° and the track steady-state error is less than 13m, which meets the standard requirements. The course rudder proposed in this paper has important reference significance for the localization of high-end navigation equipment.

In order to solve the problems of complex navigation scenes of inland waterway vessels, few shape and color characteristics of marine radar images, and the difficulty of annotation, an improved YOLOv8 marine radar image target detection method was proposed. Firstly,  to alleviate the issues of annotation errors and model overfitting, a label smoothing strategy was introduced during the model training phase. Then, combining the unique positional prior information of the radar images, a coordinate-based convolution structure was designed to simultaneously extract the shape, color and positional features of the target. To verify the effectiveness and superiority of the proposed method, comparative experiments were conducted on the collected radar images of the Yangtze River channel  under different weather conditions. Results show that the proposed method achieves an accuracy rate of 91.52% while ensuring real-time object detection, with an average accuracy  improvement of  5.17% compared to the classic YOLOv8, which can provide technical support for improving the modernization and intelligent management level of inland waterway shipping.

An improved A* and path smoothing algorithm was proposed to overcome the defects of A* based paths for unmanned boats in inland lake environments including collision avoidance insufficiency, maneuver redundancy, and steering non-smoothness. Firstly, the evaluation function was enhanced by incorporating obstacle avoidance and steering costs. Secondly, the redundant path of the A* algorithm was removed by iteratively generating the segmentations in a reversal manner. Finally, the steering trajectory was further optimized by addressing steering performance of the unmanned boat. Simulation experiments on a real inland lake map show that the improved algorithm significantly reduces the collision probability, steering angle and traveling distance, simultaneously, and thereby improving the steering performance of the unmanned boat.

In order to improve the control performance of the fault-tolerant permanent magnet vernier rim-driven motor (FTPMV-RDM) position sensorless control system, the speed controller and rotor position estimation algorithm of the system were studied. A fast super-twisting sliding mode controller was used to reduce the chattering phenomenon existing in traditional sliding mode controller, and an adaptive sliding mode observer was used to estimate the rotor position to improve the position estimation accuracy. A simulation model of the FTPMV-RDM position sensorless control system was built in the Matlab/Simulink environment. The simulation results show that the control method proposed in this paper can accurately estimate the speed and rotor position information of FTPMV-RDM in the absence of fault and one-phase open-circuit fault, and the system has good dynamic and static performance.

Red tide is a common marine ecological disaster caused by the explosive proliferation or high aggregation of certain phytoplankton, protozoa, or bacteria in seawater under specific environmental conditions, leading to discoloration of the water. At present, the main methods for monitoring red tide include fixed-point monitoring of ships, monitoring stations, and aerial and satellite remote sensing monitoring. Among them, satellite remote sensing monitoring is widely used due to its wide monitoring range and comprehensive monitoring information. This article proposes a linear comprehensive identification index (LCI) method for red tide identification in multispectral satellite images based on five sub indices and their comprehensive methods that can indicate red tide outbreaks. Based on the multi day continuous monitoring spectral data of red tide in Xinghai Bay, Dalian City, support vector machine and logistic regression classification were carried out, and the classification plane coefficients were obtained as normalized sub index combination weights. Two optimal combinations of red tide identification indices were obtained. Both methods achieved classification results with an F1 score of 0.86 or higher on the validation set, and the LCI value showed a positive correlation with the probability of red tide occurrence. This article further applies the obtained classification coefficients to the extraction of red tide pixels in multispectral satellite images, and the results show that both linear composite indices have successfully extracted red tide areas in the images. The comprehensive identification index of red tide proposed in this article provides a new spectral index calculation method for red tide identification in remote sensing images.

The fluid swivel is an important component in the FPSO single point mooring system that performs the functions of rotation, sealing and fluid transmission. To verify the safety and reliability of the function and structure of the fluid swivel in the long-term marine environment, this paper conducts research on the comprehensive performance test method of the fluid swivel. According to the performance index requirements of the newly added 8-inch domestic fluid swivel products of the Nanhai-Fenjin FPSO, the test device for simulating the dynamic marine environment was designed, and 2 running-in tests and 4 functional pressure tests were formulated. Through the above tests, the sealing pressure, rotational torque and fluid transmission performance of the fluid swivel were tested under static and dynamic swing conditions. The test results show that the comprehensive performance of the domestic fluid swivel meets the design requirements, which provides ideas and guidance for future performance testing and verification of the domestic fluid swivel. 

Dual-cycling operation is an effective way to improve container handling efficiency and resource utilization of container terminal. The complexity of quay crane scheduling with dual-cycling operation lies in the need to make decisions on the sequence of container handing operations, with considering the constraints of the ship's stability, heel angle, and hatch cover. An optimization model for quay crane scheduling is established, which takes the minimum energy consumption of quay crane as the optimization target. The improved variable neighborhood search algorithm based on the Johnson rule is designed to solve the model. The validity of the model and algorithm was verified, highlighting the importance of constraining the ship's stability and heel angle during the handing operation. It has been determined that the optimal time interval for checking heel angle is 2 min. The results showed that the model proposed can improve the handing efficiency and reduce the energy consumption of quay crane, all while ensuring the safety of container handing operation. 

In response to the issue of deviation in the meandering motion of shipboard fire snake-like robots, which hinders their effective movement within the complex gaps of vessels, based on the kinematic model of the hyper-redundant snake-like robot, the nonlinear offset causes are analyzed from the kinematic and dynamic perspectives, which are caused by the torsional torque generated during the rotation process between the joints. The experimental results show the influence rules of the swing amplitude A and the phase difference β between the adjacent links on the normal offset of the robot, proving that the two parameters can only affect the degree of the normal offset of the robot, and cannot effectively eliminate the offset. Based on the Serpenoid control curve, a simple and efficient adaptive offset correction control algorithm is established, and Webots robot simulation and prototype experiments are carried out. The results show that the average maximum radial offset of the fire-fighting snake-like robot is reduced by 42.8cm, decreasing by 75.2%, meeting the work requirements of the narrow environment of the ship without changing the shape of the robot motion.

How to ensure the safe evacuation of personnel in port accidents has become a key research topic in port planning and operation. The key to solving the problem of crowd evacuation is to understand the movement patterns of the crowd. The key to solve the problem of crowd evacuation is to understand the movement law of the crowd. Crowd movement is a complex system, involving many factors such as crowd interaction behavior, urban spatial form, and architectural environ-ment. In order to provide accurate and reliable research basis for the study of crowd evacuation, this paper proposes a crowd motion simulation model based on deep convolutional neural network. In order to obtain the data required for neural net-work training, this paper uses CSRNet neural network and DBSCAN algorithm to extract real crowd trajectory data from surveillance video. Through the training of deep convolutional neural network, the real crowd behavior pattern is deeply studied, and the crowd motion simulation model is established by using the trained deep convolutional neural network. The experimental results show that the model can accurately predict the movement behavior of the crowd and truly simulate the movement trajectory of the crowd, which can provide a basis for the formulation of emergency evacuation strategies and the design of evacuation channels in public places.

Aiming at the problem of intelligent ship trajectory tracking control under network spoofing attacks, an improved reaching law sliding mode trajectory tracking control scheme with observer is proposed. Using Bernoulli distribution to generate random binary numbers to simulate network spoofing attacks. By designing observers to observe the impact of attacks and feedback compensation to the controller to resist the damage caused by attacks on the system. Design adaptive laws to cope with the impact of environmental disturbances, while further enhancing system robustness through coordination with the observer. Design an improved sliding mode reaching law with fast compensation for large errors and smooth compensation for small errors, which reduces the chattering problem of traditional sliding mode control and improves the control efficiency of the controller. Finally, the overall stability of the system was verified using Lyapunov stability theory, and comparative simulation experiments were conducted using Matlab. The experimental results show that the control scheme can respond quickly to the impact of attacks and provide good compensation, and the trajectory tracking error can quickly converge to a very small bounded interval near zero, achieving ship trajectory tracking in a network attack environment. This plan can provide certain reference value for research on safe navigation of ships and military confrontation of maritime equipment.

Under the impetus of the International Maritime Organization (IMO) emission reduction strategies and government policies to reduce emissions, shipping companies have started retrofitting vessels to use green fuels. However, the retrofitting of vessels disrupts regular shipping operations, indirectly impeding vessel retrofitting plans. In this context, it is proposed that shipping companies cooperate through fleet alliances, coordinating and scheduling transportation tasks to increase the rate of vessel retrofitting. To address these issues, a two-level optimization model is established. In the upper level, the objective is to maximize the overall economic benefits of the alliance. In the lower level, the focus is on maximizing the profits of liner shipping operations based on the retrofitting plans from the upper level. This model represents a multi-leader-single-follower game involving three alliance companies and liner shipping tasks, resulting in an equilibrium problem with equilibrium constraints (EPEC). Using Karush-Kuhn-Tucker (KKT) conditions, the model is transformed into multiple mathematical programs with equilibrium constraints (MPEC), and a diagonalization algorithm (DM) is applied for solving. Sensitivity analyses are conducted on the alliance effects. The results show that this model and its solving method are feasible and can provide new insights for decarbonization in the shipping industry.

Considering the asymmetry of market size between shipping companies in the shipping service supply chain, by using Stackelberg game theory, a three-stage sequential game model between shipping companies and freight forwarders was established to determine the optimal encroachment and pricing strategies of shipping companies, and the impact of market share on shipping company encroaching decisions and supply chain member pricing decisions was studied. Results show that whether shipping companies encroach or not, the higher market share always brings more profits to shipping companies; when the market share of the shipping company is high, the substitution degree of transportation services between shipping companies is low, and the encroachment cost is high, the shipping company will gain more profits without encroachment; when the shippers have a higher or lower preference for encroaching channel, the encroachment will bring more benefits to the shipping company.

Aiming at an innovative engineering structure spanning the bay, the submerged floating tunnel (SFT), the three-dimensional solid model of SFT was established by using ABAQUS finite element analysis software, and taking the SFT design scheme of Funkawan Bay in Japan as the object. The lateral deformation and torsional motion response characteristics of the SFT were analyzed under the conditions of single cable break and multiple cables continuous break, and the influence of Weight-Buoyancy Ratio (WBR) on cable break response was discussed through parameter sensitivity analysis. The results show that the maximum torsional angle response and vertical displacement response of two consecutive anchor cable failures at the mid span increase by nearly three times and double respectively compared to a single anchor cable failure. Increasing the safety factor of the cable is significantly important to avoid continuous cable-breaking. In addition, the SFT equivalent model for dynamic responses analysis is proposed, which can accurately reflect the dynamic characteristics of the whole tunnel and the deformation characteristics under various loads.

A lightweight stacked residual low-light image enhancement network is proposed to overcome the difficulty in accurately sensing the environment under low-light conditions for Unmanned surface vessels (USV). Firstly, a pyramid multi-scale pooling is introduced into feature fusion to better preserve image details. Secondly, depthwise separable convolution is introduced to lighten the network and improve the image processing speed. Thirdly, a new composite loss function with color loss is designed to reduce color distortion. Finally, LeakyReLU activation function is used to prevent neuronal death. Results from extensive experiments demonstrate that our method can effectively improve image quality while speeding up image processing compared to Stacked Attention Residual Network (SARN). Specifically, SSIM (Structure Similarity Index Measure) and PSNR (Peak Signal-to-Noise Ratio) are improved by 3.31% and 2.08% and FLOPs (Floating point operations), params and processing time of single image are reduced by 81.88%, 75% and 43.02%. Therefore, the proposed method is suitable for low-light image enhancement of USV, and has certain application value.

In the context of a bus fleet combining gasoline and electric vehicles, the introduction of on-route wireless charging infrastructure significantly influences their operation and scheduling. This study revolves around the challenge of optimizing the placement of wireless charging stations and formulating operational strategies for such a hybrid bus fleet. A mixed-integer programming model has been thoughtfully devised, taking into account an array of constraints, including variables related to bus operations and the state of electric bus batteries. The objective is to minimize the overall operational costs. To tackle this model, an adaptive large neighborhood search algorithm is employed. Additionally, practical scenarios are meticulously designed to simulate a variety of fleet compositions and charging strategies. A comprehensive comparative analysis of overall operational costs is undertaken, accounting for different fleet compositions and charging strategies. Furthermore, a sensitivity analysis is conducted to scrutinize the effects of the charging power of wireless charging facilities on the results. The results underscore that in an environment featuring on-route wireless charging, the hybrid bus fleet stands out with the lowest total operational costs. Moreover, with the increase in charging power of wireless charging facilities, there is a concurrent reduction in the number of planned charging stations and gasoline-fueled buses, culminating in reduced overall costs. Therefore, the proposed model in this study offers an effective and efficient approach for significantly trimming down the overall operational costs of the bus fleet in the context of on-route wireless charging.

To cope with the impact of frequent yard crane failures in the import container yard of a container terminal and to improve the efficiency of the yard’s container lifting operations, this paper investigates the joint optimization problem of yard crane maintenance and scheduling. To be specific, a mathematical model that focuses on the impact of yard crane operation volume on its failure probability is proposed by considering the safety interval of yard crane operation and the waiting time limit of container trucks. This model aims to minimize operation, maintenance, and opportunity costs due to failure. To solve the model, a genetic algorithm embedded in a large neighborhood search is designed in accordance with the model characteristics. Numerical experiments are conducted on the imported container yard of a terminal in Dalian port to verify the effectiveness of the proposed scheme and algorithm. Compared with other scheduling schemes and maintenance strategies, the proposed scheduling scheme reduces the average cost by 3.35% and 2.84%, respectively. The research results can provide decision support for the daily operation of the terminal.

In order to investigate the influence of the stator vane passage shape on the internal flow field of a centrifugal compressor, this study focuses on the stator vane of a high-pressure centrifugal compressor. The study employs a method of middle arc line shaping, parametric modeling, and utilizes high-precision surrogate models to expedite the design optimization process. Data mining tools are used to analyze the sensitivity of design parameters and summarize the general design principles for stator vanes. The results show a close correlation between vane shape parameters and isentropic efficiency, with a moderate positive correlation, and strong interrelationships among various shaping variables. Some shaping parameters have a positive effect on static pressure efficiency. The optimal shaping structure indicates that reducing the inlet flow angle of the stator vane can alleviate the formation of separation vortices on the suction side and decreasing the outlet flow angle of the stator vane can reduce flow losses. Increasing the curvature from the middle to the trailing edge of the vane grid can mitigate blockage in the middle of the passage, leading to overall improvements in the performance of the centrifugal compressor's stator vane passage. After shape optimization, the isentropic efficiency at the design point increased by 1.7%, and the pressure ratio increased by 0.1 across the entire operating range.

In order to effectively utilize the large amount of boil off gas (BOG) generated during the operation of liquefied natural gas (LNG) ships, the ship allocation problem for LNG sea transportation routes considering BOG management was studied. A ship allocation model was established for LNG sea transportation routes at economical speeds using BOG generated from LNG as propulsion fuel. On this basis, a BOG management strategy was designed based on the relationship between ship speed and BOG demand, while a joint optimization model for LNG sea transportation route allocation and ship speed considering BOG management was established. Taking an LNG shipping company in China as an example, the lowest cost, optimal ship allocation plan, and voyage schedule corresponding to two models were obtained. Comparing the optimal solutions of the two models, it can be seen that the BOG management strategy can effectively reduce costs by more than 8%, while corresponding management suggestions are proposed from the perspective of LNG spot price fluctuations and the business scale of shipping companies. Research results show that the proposed method can provide decision support for LNG shipping companies to reasonably allocate transportation resources and effectively manage LNG ship evaporation gas.

A finite control set model predictive control strategy based on the improved virtual space vector (IVSV-FCSMPC) was proposed for DC side midpoint potential imbalance problem of midpoint-clamped three-level rectifiers. Firstly, based on multi-vector synthesis and the characteristic that virtual space vectors do not cause midpoint potential fluctuations,a single-objective value function based on current tracking control was constructed to achieve unweighted coefficient based predictive control. Secondly, a dynamic adjustment factor based on midpoint potential feedback was designed to adjust the action time of positive and negative small vectors and medium vectors in the vector synthesis process. Two pairs of redundant small vectors were simultaneously adjusted within the small sector range of the low modulation regime. In the small sector range of the high modulation regime without redundant vector action, the virtual space vector switch sequence was optimized by improving the virtual medium vector to suppress midpoint potential fluctuations, achieving precise control of midpoint potential balance within the full modulation regime range. Finally, the experimental verification was carried out based on the hardware platform of the three-level rectifier. Results show that the designed control method has good current tracking control performance, excellent midpoint potential balance control and fast response speed.

To investigate the influence of different pressure relief port areas on the evolution of hydrogen-air flame morphology, a self-built experimental platform was used for hydrogen-air explosions experiments , and OpenFOAM open-source software was used for numerical simulation to analyze the detonation characteristics and flame morphology under different pressure relief port areas (0%, 25%, 50%, 75%, 100%). Results show that the area of the pressure relief port has little effect on the flame shape, but it affects the timing of flame formation. As the area of the pressure relief port increases, the propagation speed of the tulip-flame increases, the detonation pressure decreases, and the peak appearance time is delayed. The change in pressure relief port area affects the flow field structure of unburned gas and thus affects the flame shape. A smaller pressure relief port area is conducive to the formation of vortex structures in unburned gas, promoting the formation of flat plate flames and tulip flames. The area of the pressure relief port  is an important factor affecting the shape and detonation characteristics of hydrogen flames. By optimizing the structural parameters of the pressure relief port, the propagation speed  and detonation pressure of hydrogen flames can be effectively controlled, reducing the risk of explosion and having significant implications for the safe use of  hydrogen energy.

In order to allocate yard storage space reasonably, reduce waiting time for trucks at the yard bridges, and improve yard operation efficiency, the impact of uneven container volume in the yard area on the efficiency of bridge crane loading and unloading operations was considered in this study. A mathematical model for yard space allocation with the objective of minimizing the sum of queueing time cost and driving time cost for trucks was established. For the cost of queueing time for trucks, an embedded truck operation simulation model based on Anylogic was used to model the mathematical relationship between average queueing time at the yard bridges, bridge crane loading and unloading efficiency, and truck arrival rate. A variable neighborhood search algorithm was designed to solve the model for large-scale instances, and a series of numerical experiments were conducted to evaluate the performance of the proposed algorithm. The study investigated the degree of impact of changes in yard loading and unloading efficiency on yard space allocation decisions. The results indicates that considering the impact of uneven container volume in the yard area on bridge crane loading and unloading efficiency can effectively reduce the queueing time cost for trucks waiting for bridge crane operations (by approximately 14.74%) and accelerate truck operation turnover, thereby improving yard operation efficiency.

To optimize the trajectory planning problem of underactuated ship automatic berthing and avoid getting stuck in local optima, a real-time trajectory planning method considering the overall constraints of the ship kinematic model and obstacle constraints was proposed. Based on the gridded map, the A* algorithm was used to plan the global optimal path, and the redundant point deletion strategy was applied to obtain the reference waypoints for trajectory planning. Considering that ship berthing was a dynamic process, the extended dynamic window algorithm (EDWA) was used to predict the berthing trajectory of the ship during the constant force stage and deceleration stage, and the proportional integral derivative (PID) control method was introduced to adjust the ship’s heading. After conducting feasibility tests on all predicted trajectories, the optimal trajectory was obtained by using an evaluation function that considers constraints such as target points, reference waypoints, and obstacles, and the corresponding force was used as the control input for the ship. Repeat the above steps for realtime planning based on the current control input and motion state of the ship at the next sampling time until the ship reached the target point. The feasibility of the proposed method was verified by using the simulation scenario of Rizhao port. Results show that the proposed method can guide ships to avoid obstacles and reach the designated berth in a relatively short time, the distance between the trajectory and the nearest obstacle is 144.1428 m, and the trajectory length is 47.26 m longer than the traditional A*+EDWA, however, the planning time is reduced by 55.12%, which show the superiority of the proposed method.

The historical intensity method and benchmark value method are commonly used methods for initial allocation of carbon quotas.At present, only Shanghai and Shenzhen in China have included the port industry in the carbon trading system, and both use the historical intensity method for initial allocation of quotas.This paper proposes five carbon emission quota allocation criteria for the port industry based on the principles of efficiency and fairness, namely optimal industry emission reduction efficiency, optimal carbon emission efficiency, encouragement of advanced players, historical emission responsibility, and large enterprise responsibility, and constructs a comprehensive evaluation model for carbon quota allocation in the port industry based on TOPSIS entropy weight method.Empirical analysis is used to evaluate the efficiency and fairness of three quota allocation methods: historical intensity method, benchmark value method based on carbon intensity ranking, and mixed allocation method based on energy consumption limit.The results show that under the efficiency-fairness integrated objective and the two sub-objectives of efficiency and fairness, the relative proximity between the evaluation index and the optimal solution of the mixed allocation method based on the energy consumption limit is 0.566, 0.711 and 0.638, respectively, ranking first among the three allocation methods and obviously higher than the other two methods under the fair sub-target and comprehensive evaluation target.Although the mixed distribution method increases the complexity of the specific implementation process, it also improves the efficiency and fairness of the distribution results. 

A self-following dual-mode obstacle avoidance model predictive control method based on improved dynamic window method was proposed for the trajectory tracking control problem of unmanned surface vehicle (USV) formation under the influence of sudden obstacles. Firstly, the longitudinal velocity and bow angular velocity oscillation constraints were introduced to reduce the jitter of the velocity and bow angle in the design of the evaluation function of the dynamic window algorithm so as to achieve better integration of obstacle avoidance planning and control. A strategy of calculating the steering azimuth based on intermediate distance was proposed to reduce the sharp turns in the obstacle avoidance process of the USV, while an evaluation term based on the deviation of the predicted and the expected position ending was designed. Meanwhile, the formation keeping information was combined to correct the obstacle avoidance endpoint and reduce the deviation of the USV formation. Secondly, based on the linearized Taylor expansion principle, a prediction model for USV formation was designed. The prediction value of the model was corrected by the prediction error between the system output measurement and the model prediction values. At the same time, a rolling finite time domain iterative online optimization strategy was adopted to propose a trajectory tracking model prediction control method for USV formation that integrated autonomous following dual-mode obstacle avoidance strategy. Finally, a nonlinear disturbance observer was designed to compensate the environmental disturbances, while the Lyapunov function was constructed to prove the stability of the system by combining with the terminal penalty theory. Eventually, the effectiveness and reliability of the proposed USV formation obstacle avoidance and trajectory tracking control algorithm were verified by simulation experiments.

Aiming at the rock instability caused by hydro(H)-machine(M)-damage(D) coupling during the construction process of foundation pit engineering in karst area, the stress-seepage-damage coupling model and algorithm based on Hoek-Brown strength criterion are established. Firstly, according to the damage theory, the H-B strength criterion was modified by introducing damage factor, and the governing equations of stress field and seepage field and permeability evolution equation were given to realize the complete coupling of stress, seepage and damage. Then, a solution algorithm for the model was provided, and the results of triaxial indoor tests and simulation calculations of limestone were compared, which showed good fitting and verified the accuracy of the model. Finally, the model was used for the excavation stability analysis of the foundation pit project of the Suoyuwan South Station of Dalian Metro, and the surface settlement and underground continuous wall deformation under the coupling and non-coupling conditions were calculated respectively. The calculation results show that the deformation value considering the coupling is significantly greater than the uncoupling, and it is closer to the field monitoring value. At the same time, in order to determine the minimum thickness of the anti-outburst layer of the surrounding rock, the distribution characteristics of the damage area at 2m, 4m, 6m and 8m of the anti-outburst layer were calculated, which has certain significance for the risk prediction during actual construction.

Aiming at the problem of poor performance of traditional predictive function control (PFC) in speed control due to external disturbances, parameter mismatches, and measurement and control quantization errors caused by position sensors in the servo control system of permanent magnet synchronous motor (PMSM), an optimized predictive function control method was used by using a composite adaptive position observer. The composite adaptive observer was constructed by combining model reference adaptive control (MRAC)  with extended state observer (ESO) to improve the locationless observer observation accuracy and disturbance resistance, accurately detected rotor information, and used it as feedforward compensation for the predictive function controller, which achieved fast and high performance control of the overall servo system. The experimental results show that the improved  predictive function control has feasibility and superiority in response speed and disturbance suppression under sudden changes in load and speed.

A dual-branch residual convolutional neural network was proposed for image enhancement in PIV velocity technique to obtain high-quality particle images. Firstly, a dual-branch convolutional neural network composed of residual blocks was designed to extract features from the input particle image pairs, while a coding-decoder was used to effectively fuse the feature information of the particle image pairs. Secondly, a challenging image enhancement dataset was autonomously generated to train model parameters, including Gaussian noises of different concentrations, light intensity noise and various real interference backgrounds, thereby fully simulating real fluid scenes. Results show that the proposed method can effectively deal with noise interference in both synthesized and real images, achieving image enhancement. Meanwhile， higher precision velocity fields can be obtained by using velocity field estimation algorithm to process the particle image pairs enhanced by the proposed method in this paper.

In order to study the influence of panic emotion transmission among passenger ship personnel, a P-SEIRS panic emotion infection model was established based on the personality OCEAN model and SEIRS infectious disease model. A personnel evacuation model considering panic emotion infection was established by using Anylogic evacuation simulation software, the evacuation process of passenger ship personnel under panic emotion transmission was analyzed, and the influence of initial panic personnel proportion, sensitivity, equipment intervention, and number of guides on panic emotion transmission and evacuation efficiency was explored. The simulation results show that panicked individuals will exacerbate congestion, leading to an increase in evacuation time; the higher the sensitivity of the crowd, the faster the spread of panic emotions, and more panicked pedestrians will be generated during the evacuation process; equipment intervention and appropriate guidance are beneficial for improving passenger evacuation efficiency, but excessive guidance can prolong evacuation time.

In order to explore the influence of factors such as water content of foamed asphalt, cement/lime, asphalt aging degree and temperature on the interface characteristics of foamed asphalt cold recycling binder (composed of cement/lime, aging asphalt and foamed asphalt, hereinafter referred to as binder), molecular dynamics method was used to build different types of binder molecular models, and the interface energy, interaction energy, cohesive energy density (CED) and water molecule radial distribution function (RDF) were used as the calculation indicators to quantitatively analyze the interface characteristics of each component material in the binder, thus revealing the microscopic mechanism of interaction between each component material. Results show that, in terms of the same material properties of other components, the binder with long-term aging asphalt (LTAA) exhibits lower interfacial energy, interaction energy, CED and water molecules RDF compared to binder with short-term aging asphalt (STAA). When the moisture content of foamed asphalt in binder is 2.4%, the interfacial energy, interaction energy, CED and water molecular RDF of binder are the highest, and when the moisture content of foamed asphalt in binder is 1.8%, the interfacial energy, interaction energy, CED and water molecular RDF of binder are the lowest. Compared with lime characterized by CH, cement binder characterized by CSH has higher interfacial energy and interaction energy, but its CED and water molecule RDF are lower.  When the temperature is 303 K~308 K, the binder containing STAA has higher interface energy and interaction energy than the binder containing LTAA, but that is opposite when the temperature is 313 K~318 K, and the CED of the binder basically remains stabilization as the temperature increases. Compared with the cementing material containing CH, the cementing material containing CSH has higher interfacial energy and interaction energy, which gradually increases with the increase of temperature and reach the highest value at 318 K, however, the CED of the both remains stabilization. As the temperature increases, the RDF of water molecules in different binders also increases, and a large number of water molecules concentrate towards the ends, which has a negative impact on the interfacial characteristics of binders.

In order to improve the speed and accuracy of water level prediction, a prediction model based on improved echo state network (ESN) was proposed. Xavier method was introduced to optimize the weights to adapt to the water level prediction task. At the same time, the concept drift detection method (EDDM) was introduced to adapt to the actual water level environment, monitor the change of water level data distribution, and trigger the corresponding model update or adaptation strategy when the concept drift was detected, so as to improve the prediction effect of the real water level. The experimental results of data from nine water level stations show that compared to traditional sequence prediction models (SVM, RNN, GRU, LSTM, ESN, and XESN(Xavier-ESN)), the proposed model shows higher prediction accuracy for each water level station in terms of overall prediction performance and short -term, medium - term and long-term predictions, further improving the prediction accuracy of  inland waterway water level.

In order to improve loading efficiency, focusing on the pre-marshalling problem and taking the priority of the slot-storage export containers as the main decision variable, a priority model for the pre-marshalling problem of export containers in storage yard was proposed. Considering containers flipping operations as the core objective of the decision process, a transcoding technique was introduced based on the properties of decimal and binary conversion. Combining with existing research models and the proposed integer programming model, a set of transcoding models were constructed. By testing on existing data sets, compared with existing models, the proposed priority model can reduce the average solution time by 75.66%,while the solution time of the transcoding model group is significantly lower than that of the original model group, in addition that the maximum solution efficiency is increased by 182.5%.

In order to explore the influence of cavity leakage flow on the aerodynamic performance and flow field structure of the annular diffuser cascade, the annular diffuser cascade was studied based on the verified numerical simulation method, and the effects of leakage flow rate and circumferential velocity on corner separation and total pressure loss were analyzed. Results show that as the leakage flow increases, the expansion range of the passage vortex (PV) becomes larger, the corner separation occurs earlier, and the total pressure loss coefficient of the blade gradually increases.When the leakage flow coefficient increases to 1.5%, the total pressure loss coefficient expands by 22.9% . The increase in circumferential velocity of the leakage flow improves the blockage situation in the near end wall area.The degree of deflection of the fluid close to the end wall towards the suction surface is weakened, and the range of deflection  is reduced. When the circumferential velocity coefficient increases to 1.1, the total pressure loss  coefficient  decreases by 21.3%.

With the rapid development of science and technology, multi-unmanned ship systems have shown great potential in military, rescue and escort mission scenarios. The purpose of this paper is to explore the formation construction problem of multiple unmanned surface vehicle systems based on multi-agent deep reinforcement learning algorithm. Considering the sluggish convergence speed of the conventional multi-agent deep deterministic policy gradient algorithm (MADDPG), this study incorporates the attention mechanism into the value function stage to enhance the convergence speed of the formation decision model for a multi-UAV system. Through the cooperation of the formation model of the unmanned surface vehicle with the formation collision avoidance and the formation construction reward function, the efficiency of the multi-UAV to complete the formation construction task is finally improved. The simulation results conclusively demonstrate the efficacy of the proposed method in accomplishing multi-unmanned surface vehicle formation construction tasks, thereby establishing a solid theoretical foundation for future applications of multi-unmanned ship formation construction.

Focusing on the core audience of the Mobility as a Service （MaaS）service platform, namely public transportation users, this study aimed to investigate the underlying mechanisms of their willingness to use the MaaS service platform and gained a deeper understanding of the driving factors that encouraged this group to use MaaS services. This paper divided the public transportation user group into passive passenger group and selective passenger group. Based on the questionnaire survey data of Dalian City, Liaoning Province, a multi-index and multi-factor structural equation model was used to deeply explore the satisfaction level of these two groups with the current level of public transportation services, and their attitudes and behavioral intentions toward using MaaS services. Research has found that passengers’ attitudes and willingness to use MaaS services are positively correlated with their satisfaction with public transportation services, but there are mechanism differences between the two types of passengers in the process of converting satisfaction into MaaS usage willingness. Selective passengers’ satisfaction directly converts into usage intention, while passive passengers first change their attitudes towards MaaS, thereby affecting their usage behavior. Therefore, when promoting MaaS services, differentiated strategies should be developed for different passenger groups，that is，for selective passengers, the efficiency and convenience of MaaS services should be highlighted, while for passive passengers, the quality of MaaS services and environmental comfort should be reflected. In addition, factors such as gender, age, and transfer frequency also affect the willingness to use MaaS. The above findings provide decision-making basis for relevant departments to plan MaaS service platforms, optimize public transportation services, and improve MaaS penetration rates.

When optimizing the reservation quota scheme, existing methods faced many challenges, including difficulty in capturing dynamics and randomness, limited theoretical assumptions, and complex data acquisition and processing. GPS trajectory data of container trucks was used to comprehensively capture the entire process of container truck turnover from arrival at the port to completion of operations, and based on this, establishes a mapping relationship between the number of container truck arrivals within the unit appointment time window and their total turnover time at the port. A more precise and concise quota optimization model was constructed by using this mapping relationship, and the quota appointment system for container trucks was optimized and solved. The experimental results show that the optimized reservation quota scheme effectively reduces the arrival volume of container trucks during peak periods, while increasing the number of vehicles arriving during off peak periods, significantly reducing the total turnover time of container trucks at the port, improving the operational efficiency of the terminal, and thereby enhancing the competitiveness of the port.