Efficient reconstruction of three-dimensional multiphysics fields—including temperature, velocity, species concentration, and pollutant formation—is essential for monitoring the operational status of energy systems. However, direct measurements are typically sparse, while high-fidelity Computational Fluid Dynamics (CFD) simulations remain computationally prohibitive. To address these challenges, this study proposes a physics-informed subdomain Proper Orthogonal Decomposition (POD) framework for the rapid and accurate reconstruction of multiphysics fields in industrial gas-fired boilers. First, a surrogate modeling strategy is employed to map operating parameters to POD coefficients. Six surrogate models, including artificial neural interpolation, Gaussian process regression, polynomial regression, radial basis function interpolation, random forests, and support vector regression, are systematically evaluated. Among them, the random forest model is selected due to its superior predictive accuracy and robustness. Second, a reduced-order modeling approach integrates POD-based dimensionality reduction with surrogate models, enabling field reconstruction within seconds. Third, a physics-informed subdomain method partitions the combustion domain according to turbulence intensity and adaptively determines the optimal number of POD modes based on local energy thresholds. Validation on a gas-fired boiler demonstrates that the proposed framework achieves high reconstruction accuracy, outperforming both global and uniformly partitioned approaches. Overall, this framework provides a practical engineering solution that significantly enhances computational efficiency while further reducing reconstruction errors.

This paper proposes a novel particle swarmsimulated annealing (PSO-SA) algorithm for ship global route planning. This algorithm modifies the inertia weight in the traditional PSO to a sinusoidal adaptive form and integrates a simulated annealing approach to address the issue of local convergence, thereby obtaining the optimal route for ship routing. The results demonstrate that the PSO-SA algorithm enhances the swarm's ability to escape local optima during the search process and improves the algorithm's convergence accuracy. Additionally, by taking into account the ships' maneuverability, the algorithm optimizes a global route that aligns with nautical practice and exhibits strong applicability.

The tracking of the ship trajectory plays a critical role in the monitoring of maritime safety and navigation optimization. As a representative of multi-model target-tracking algorithms, the Interacting Multiple Model (IMM) algorithm achieves efficient trajectory tracking through the interaction of multiple filters, where the dynamic calculation of model weights is crucial. Based on the traditional IMM algorithm, this paper integrates Bayesian network technology to construct a dynamic model of ship motion. Through Bayesian inference, the IMM model weights are adaptively adjusted according to the probability distribution of the ship's motion attitude nodes, thereby enhancing the accuracy of state estimation. Experimental results demonstrate that the proposed method significantly enhances the adaptability and accuracy of the IMM algorithm in complex maritime environments, providing critical support for enhancing the reliability and robustness of ship trajectory tracking.

Accurate long-range forecasting of the El -Southern Oscillation (ENSO) is vital for global climate prediction and disaster risk management. Yet, limited understanding of ENSO's physical mechanisms constrains both numerical and deep learning approaches, which often struggle to balance predictive accuracy with physical interpretability. Here, we introduce a data driven model for ENSO prediction based on conditional diffusion model. By constructing a probabilistic mapping from historical to future states using higher-order Markov chain, our model explicitly quantifies intrinsic uncertainty. The approach achieves extending lead times of state-of-the-art methods, resolving early development signals of the spring predictability barrier, and faithfully reproducing the spatiotemporal evolution of historical extreme events. The most striking implication is that our analysis reveals that the reverse diffusion process inherently encodes the classical recharge-discharge mechanism, with its operational dynamics exhibiting remarkable consistency with the governing principles of the van der Pol oscillator equation. These findings establish diffusion models as a new paradigm for ENSO forecasting, offering not only superior probabilistic skill but also a physically grounded theoretical framework that bridges data-driven prediction with deterministic dynamical systems, thereby advancing the study of complex geophysical processes.

A high-performance heading controller plays a crucial role in navigational safety. Enhancing the controller’s anti-disturbance capability and response speed often introduce a trade-off with increased control jitters. Therefore, it is essential to improve the overall performance. In this study, we propose the ship heading controller named BESO-PPF that introduces backstepping control and the extended state observer (ESO) with the PPF, which stands for the particle swarm optimization (PSO), the parameter adjustment index (PAI), and the fuzzy rules, to optimize the system parameters. Initially, the controller, which combines backstepping control with an improved ESO with higher accuracy, is designed to enhance the system’s anti-disturbance ability. The stability of this controller is proven using the Lyapunov method. Then, the response speed of this controller is enhanced by the PSO, and the rudder jitters are mitigated through the PAI, which is optimized by the fuzzy rules. Finally, the simulation of a training vessel named Yulong is performed to validate the effectiveness of the algorithm. The results demonstrate that the performance of the optimized controller is improved significantly.

Accurate sensor network prediction is crucial for improving industrial boiler efficiency and safety. While existing predictive models show promise, they are constrained by several limitations: (i) insufficient integration of interpretable multi-level spatiotemporal information, (ii) over-reliance on static topologies and shallow features, and (iii) limited continuity and adaptability in complex environments. To address these challenges, we propose a novel framework to improve parameter prediction in gas-fired boiler systems through node similarity-based feature aggregation. First, we apply a node similarity-based multi-level aggregation strategy for interpretable multi-scale integration. Next, dynamic graph learning, utilizing a higher-order graph convolutional network, captures the evolving relationships between sensors and time steps. Additionally, continuous modeling is facilitated by a spatiotemporal ordinary differential equation solver, which overcomes the limitations of discretized time steps. Real-world evaluations show our approach improves accuracy and robustness, even with sensor failures. Furthermore, the continuous model supports predictions at any time step. This approach provides a foundation for data-driven parameter prediction and the modeling of interacting industrial components.

For ship autopilot, this paper proposes a sliding mode controller for ship heading control using a minimal learning parameter approach that combines neural networks and fuzzy methods. The system dynamics and the gain of rudder are respectively approximated by a Radial Basis Function (RBF) neural networ. The weights of the RBF network are normalized by their weight norms which reduces the learning parameters to two. In the design of sliding mode controller, a fuzzy method is introduced to dynamically adjust the parameters of the sliding mode to meet the requirements of ship maneuverability. The effectiveness of the proposed algorithm is verified through simulations using the training ship "Yu Long."

In the military field, accurately tracking an aircraft's trajectory is crucial for mission success and flight safety. However, during flight, an aircraft undergoes complex dynamic variations, including various motion patterns such as constant velocity, acceleration, deceleration, and turning. Traditional single-model approaches struggle to effectively capture these complex motion characteristics. Therefore, the Interactive Multiple Model (IMM) method has emerged as an effective tool to address this issue. IMM integrates multiple motion models and dynamically switches between them based on a probabilistic transition matrix, thus improving the accuracy of tracking complex flight maneuvers. However, the performance of IMM largely depends on the design of the prior transition probability matrix. The traditional IMM generally specifies the probability transition matrix based on empirical values commonly used by practitioners. To address this issue, this study introduces a Bayesian optimization algorithm to optimize the probability transition matrix for enhanced trajectory tracking performance. By adjusting matrix parameters, this approach significantly enhances the accuracy and robustness of the trajectory tracking. The simulation results indicate that this method can more precisely capture the actual trajectory of the motion of an aircraft during flight, achieving high tracking accuracy.

This paper proposes a fuzzy adaptive sliding mode controller to improve the quality of ship course control through system parameter optimization. First, a sliding mode function is constructed, incorporating adaptive and fuzzy control to approximate the system's switching and uncertain components, respectively. Subsequently, an ergodic method is employed to search for the optimal parameters of the Gaussian function within the fuzzy system according to the designed indicator. Finally, to enhance the interpretability of the controller, six fuzzy rules extracted from human experiences are devised to optimize the parameters of the sliding mode function. The stability of the controller is proved through Lyapunov analysis, and the effectiveness is verified using a training vessel named “Yulong”. The experimental results demonstrate that the proposed controller outperforms pre-optimization model in terms of response time, accuracy and reduction of jitter.

Enhancing the path planning capabilities of ships is crucial for ensuring navigation safety, saving time, and reducing energy consumption in complex maritime environments. Traditional methods, reliant on static algorithms and singular models, are frequently limited by the physical constraints of ships, such as turning radius, and struggle to adapt to the maritime environment’s variability and emergencies. The development of reinforcement learning has introduced new methods and perspectives to path planning by addressing complex environments, achieving multi-objective optimization, and enhancing autonomous learning and adaptability, significantly improving the performance and application scope. In this study, we introduce a two-stage path planning approach for large ships named MAPF–DQN, combining Manipulation-Compliant Artificial Potential Field (MAPF) with Deep Q-Network (DQN). In the first stage, we improve the reward function in DQN by integrating the artificial potential field method and use a time-varying greedy algorithm to search for paths. In the second stage, we use the nonlinear Nomoto model for path smoothing to enhance maneuverability. To validate the performance and effectiveness of the algorithm, we conducted extensive experiments using the model of “Yupeng” ship. Case studies and experimental results demonstrate that the MAPF–DQN algorithm can find paths that closely match the actual trajectory under normal environmental conditions and U-shaped obstacles. In summary, the MAPF–DQN algorithm not only enhances the efficiency of path planning for large ships, but also finds relatively safe and maneuverable routes, which are of great significance for maritime activities.

This study presents a global path planning algorithm utilizing an enhanced ant colony algorithm to overcome challenges like vulnerability to local optima, inadequate pheromone update, and suboptimal paths in traditional ant colony algorithms for global path planning of vessels. First, an improved heuristic function is introduced to overcome the problem of local optima. Subsequently, an adaptive pheromone evaporation coefficient is employed to balance the convergence of the algorithm in different stages. Finally, an adjacent node detection algorithm for obstacles is designed to ensure that the optimized path aligns with maritime practice. The simulation results confirm the proposed algorithm's effectiveness. This research provides valuable insights and practical recommendations for global path planning in maritime navigation.

新工科建设背景下,为提高航海类专业毕业生素质和职业能力,基于成果导向教育理念,分析航海类专业本科毕业生论文质量问题,建立基于OBE理念的毕业论文质量提升模型,分别从课程体系、产学融合和本科生导师制三个方面提出提高航海类专业本科毕业论文质量的改革措施.

Ensuring the reliable operation of industrial boilers with high production efficiency and low pollutant emissions remains a significant challenge due to the complex chemical and physical reactions that occur within the boiler system. To address this issue, the present research introduces a novel air proportioning methodology for opti-mized combustion in gas-fired boilers, incorporating real-world data, numerical-computational technologies, and a normalization method. Initially, an average discrepancy of 11.32% is observed between the airflow recorded by the sensors and the actual airflow in the gas-fired boiler. The optimum oxygen levels in the flue gas are deter-mined by striking a trade-off between heat release and pollutant emissions. At loads of 35%, 55%, 75%, and 95%, the recommended oxygen levels under equal-weighted conditions are 3.62%, 3.75%, 3.82%, and 3.91%, respectively. Furthermore, this research also considers the oxygen levels when there are non-equal weightings between heat release and nitrogen oxide (NOx) emissions. At 55% load, the weighting between heat release and pollutant emissions shifts from 1:1 to 1:2, necessitating an increase in the oxygen concentration from 3.75% to 4.30%. The air proportioning methodology proposed in this study offers an efficient framework for optimizing combustion in gas-fired boilers.

The steady turning motion of merchant ships is modeled according to industry specifications. However, challenges arise when motorboats are modeled. This study proposes a novel data-driven multi-block fuzzy cognitive map (FCM) model trained based on sea trials with four ship states. The optimal positions sampled by two different types of sensors were modeled using inverse variance weighting, which takes both Cartesian coordinate transformation and the constraints of the locations of the apparatuses into consideration. Combining these samplings with a scheme of multi-block FCMs, we conducted a study on data from motorboat trials. Our results closely approximate the data from motorboat trials at sea and are validated by a generated dataset of the classical model. Furthermore, we reveal the characteristics of our scheme, including the number of data blocks, the boundary of each block, and the parameters of the FCM for each block. As opposed to the classical method, the proposed scheme is insensitive to speed. This study presents a promising step toward mining modeling information from ship trials.

To solve the identification of a fuzzy cognitive map (FCM) with the unknown slope parameter, we propose a least-squares algorithm with slope parameter optimization for FCM in this paper. Firstly, all possible structures of FCM are decomposed according to the column vectors of the adjacency matrix. Subsequently, a functional index function is constructed to evaluate the optimal structures and weights of FCM. Then, multiple rounds of optimization search for slope parameter are performed. The weights of all structures are estimated from the observed sample data, and the optimal solution of this round is selected by the minimum index value. The range of the slope parameter is compressed according to the optimal solution until the performance is satisfied. Finally, the experiment results show the effectiveness of the proposed algorithm.

Research on modeling ship motions is beneficial to promote safety in the shipping industry. However, it has always been a difficult point in the representation of this complex nonlinear system. The fuzzy cognitive map (FCM) can represent nonlinear systems effectively, and its divide-and-conquer strategy simplifies the modeling process. We propose a scheme of multiblocks of FCMs for modeling the steady turning motion of ships. First, the physical variables of the maneuvering mathematical modeling group (MMG) are employed as the conceptual nodes, and the causal links between concepts are analysed to establish the structure of FCMs. Subsequently, combined with normalized data, the loss function is optimized by least squares method, and the automatically partitioned of the training data and the parameters estimation of the FCMs in each data block are completed simultaneously provided that the designated required accuracy are satisfied. Finally, the multiblocks FCMs are integrated into an overall Ship-FCMs system after verification by the testing data. Experiments are conducted based on the simulation data of ?Yulong?, which demonstrate the effectiveness of the proposed algorithm.

Ship automatic collision avoidance is the key technology of unmanned ship. Aiming at the problem of computational time of collision avoidance decision-making for multi-ship encounter situations, a novel fast decision-making method for ship’s course altering for collision avoidance is proposed. To reduce the number of calculation targets, cluster analysis is introduced to classify multiple target ships with similar attributes into one group ship. Under the constraints of International Regulations for Preventing Collisions at Sea, candidate solutions are determined based on own ship’s length and encounter situation. Taking the voyage losses as the objective function, a deterministic optimization algorithm is used to rapidly obtain the optimal course altering decision for collision avoidance in the global scope. The simulation results show that the method is safe and efficient, and can provide decision support for the automatic collision avoidance of ships.

In order to process the missing turning motion observations of training boat, a data filling model scheme based on the combination of ARIMA (Autoregressive integrated moving average model) and Kalman filtering was proposed in this paper. Firstly, ARIMA was modelled to forecast the missing course observations. Then, the ARIMA model was converted to state space model, and Kalman filtering was employed to recursively estimate the state of yaw acceleration provided the prediction accuracy is unsatisfactory. Subsequently, fuzzy rules were designed to optimize the missing observations of course to prevent errors caused by extrapolating from different initial values. Finally, the optimized series of course were utilized to estimate the missing ship positions in the polar coordinate mode. A case study of the training boat turning motion at sea verifies the effectiveness of proposed scheme, which is beneficial for the usages of sampling data and the reduction cost of ship trials.

Intelligent collision avoidance systems can increase the safety of ship navigation and reduce the influence of human error. Thus, related technologies are critical to the development of the world's shipping industry. Risk of collision is a central concept in the Convention on International Regulations for Preventing Collision at Sea, 1972 (COLREGs). Collision risk is also a key factor in determining whether a collision avoidance action is required, specifically with regard to intelligent collision avoidance decisions. On the basis of the ship domain proposed by the Fujii, at first, this paper defined the risk of collision index λ, which can be used to measure the risks between owe ship ant the target ships, then, according to the COLREGs, categorised ship encounter types as head-on situation, crossing situation, and overtaking situation, and established the action flow charts of each of these situations, at last, with the permission of Dalian Maritime University, the training ships of Yu Peng and Yu Kun were selected as simulation objects, the paper presents simulations of head-on, crossing and overtaking situations to verify the usability of the proposed models.