In order to study the influence of mass increase and the corresponding added damping and stiffness on the displacement response of beam members under a uniform blast loading, the elastic-plastic displacement solutions of beam members with mass parameters as variables were derived, using the ductility ratio to control the extent of plastic deformation. Using a rectangular section as representative case for beam members, we designed 17 calculation cases for the parameters of mass, additional damping, and additional stiffness. The effect of these parameters on the vibration displacement of beam members under uniform blast loading was analyzed, particularly the impact on peak elastic and the peak elastic-plastic displacement. The comparison of LS-DYNA numerical simulation verification and theoretical results of the displacement response of the rectangular section steel beam under blast loading was completed, which verified the reliability of the theoretical method proposed in this paper. The results show that the reduction of various peak displacements can be achieved by adding mass and the degree of reduction is lower than that of adding mass. The single effect of additional damping in the mass-damping-stiffness coupling calculation may increase the peak displacement reduction by more than 3 times of that obtained from the mass-damping coupling alone. The additional stiffness effect, when included in the mass-damping-stiffness coupling, reduces peak elastic displacements to similar degree as mass-stiffness coupling alone. The coupling calculation method of mass, additional damping, and additional stiffness should be uesd for the blast-resistant beam design.

In order to explore the regularity of dynamic response of stiffened plates under slamming load and establish static equivalent design method of stiffened plates, which covers the moving characteristics of slamming pressure and the dynamic response characteristics of stiffened plates, the slamming load acting on the flared bow is simplified in this paper into a moving pulse time history which rises linearly and decays exponentially. By using Abaqus finite element software, the elastic dynamic response of multiple stiffened plates under moving slamming load is calculated and the influence of different load parameters on the dynamic response of stiffened plates was analyzed. The Dynamic Load Factor D_LF(Dynamic stress equivalent factor) was introduced to analyze the variation rule and sensitivity of D_LF and the dimensionless parameters of slamming load. The regression and verification of D_LF formula were carried out. A parameter determination method of the simplified slamming pressure model is proposed. The comparison with the results of wedge drop test shows that the proposed slamming stress equivalentizing formula D_LF has higher precision and can effectively improve the efficiency of the preliminary design of ship structure under slamming load.

Aiming at the problem of flow noise reduction in sudden expansion pipe, this paper investigates the flow noise characteristics of viscoelastic fluid in a sudden expansion pipe and reveals the noise reduction mechanism. The flow noise in the sudden expansion pipe could be controlled by viscoelastic fluid which is obtained by mixing the surfactant with water. Based on the finite extensible nonlinear elastic model with Peterlin approximation (FENE-P), the flow noise is calculated by using Fluent UDF, combined with Large Eddy Simulation and the Ffowcs Williams-Hawkings method. The Reynolds number is 8000, the maximum polymer tensile length is 10, and the Weissenberg number (We) ranges from 0.1 to 12. The results show that when We reaches a certain value, the flow field is stabilised and the generation of vortices is delayed. Thus, the formation of large vortex that generates low-frequency noise is inhibited. The fluctuating pressure and fluctuating velocity inside the pipeline and at the pipe wall are reduced significantly, so that flow noise will be suppressed from the perspective of sound source. With the increase of We, the effect of noise reduction is enhanced and the frequency range of noise reduction is extended, but when We exceeds a certain value, the noise reduction effect reaches the threshold.

This paper is to study the principle of line spectrum feature extraction. In view of the deficiency of manual line spectrum feature extraction method, a line spectrum feature extraction method based on machine learning was proposed. The Encoder-Decoder based on convolution neural network was built, and the attention mechanism was introduced between the convolution and pooling layers, so that the important features of the input data could occupy a higher weight to enhance the accuracy of feature extraction. The model was compared with U-Net model and TPSW algorithm in the case of low signal-to-noise ratio, and tested on the actual data. The experimental results show that the improved model achieves a line positioning accuracy of 0.823 at a signal-to-noise ratio of 5 dB. This performance is better than that of the U-Net model and TPSW algorithm with in the 0~5 dB range. Thus the model effectively extracts line spectrum information and improves the accuracy of underwater target detection.

This paper focuses on a numerical algorithm for analyzing flow-induced noise of composite plates. In this study, a method based on the superposition of uncorrelated plane waves was proposed to synthesize turbulent fluctuating pressure and calculate flow-induced noise using the finite element acoustic-vibration coupling module. The detailed theory and procedures of this algorithm were provided for rectangular plates as an example. The computational results were compared with analytical theory to validate the accuracy of the algorithm. Based on the wave number filtering principle, the criteria for selecting the truncation wave number was given, and the algorithm was compared with traditional methods in terms of computational efficiency, demonstrating the advantages of the algorithm in computational efficiency and accuracy. Finally, the mechanism of flow-induced noise generation in composite plate structures with acoustic coating was analyzed in this paper using the proposed algorithm, providing theoretical support for reducing flow-induced noise of underwater vehicles.

Surface piercing propellers are one of the preferred propulsion systems for high-speed crafts due to their favorable features such as no cavitation, low resistance, shallow draft and high efficiency. Since the propellers often operate in oblique flow with blade emergence, there will be vertical forces generated which will have an important impact on the hydrodynamic performance of the planing crafts. In this paper, the effects of vertical force generated by surface piercing propellers on hydrodynamic performance of a stepped planing craft through numerical and experimental methods were studied. And the cases for planing boat with different locations of center of gravity were also involved. The results show that the vertical force produced by surface piercing propellers has a positive effect on the resistance of the planing crafts and that the effect is related to the longitudinal position of the center of gravity of the planing boat. Consequently, the longitudinal position of the center of gravity should be adjusted to the stern for planing craft equipped with surface piercing propellers, and the vertical force of the propellers should not be too large.

Accurate motion prediction is crucial for the safe navigation of Autonomous Underwater Vehicles (AUVs). A fast Least Squares-Support Vector Machine (LS-SVM) motion prediction approach based on ridge regression algorithm is proposed in this paper. Firstly, the ridge regression analysis was incorporated into the traditional hydrodynamic model and the correlation analysis was conducted on the acceleration time-series input variables to identify the relatively important components. This step effectively reduces computational complexity while maintaining prediction accuracy. Subsequently, for the issue of high dimensionality and computational complexity in the LS-SVM algorithm's kernel function matrix, an improved Lagrange function was designed to eliminate redundant bias terms. This modification lightens the burden of calculating high-dimensional kernel matrices and further enhances the speed of maneuvering prediction. Finally, case studies based on the REMUS model demonstrate that the proposed strategy, compared to the standard LS-SVM prediction method that relies on traditional hydrodynamic models, reduces computational runtime by 29.8% while ensuring prediction accuracy.

The performance of Autonomous Underwater Vehicles (AUVs) is significantly influenced by their shape design. This study presents a novel AUV shape optimization method that integrates data-driven approaches and machine learning technique to focus on the impact of the head and tail profiles, and the configuration of the fins and rudders on AUV performance. A parameterized hydrodynamic analysis workflow was developed to automatically generate the hydrodynamic parameters required during the optimization process. Based on this workflow, a data-driven framework was constructed for multi-objective optimization of AUV shapes, with the aims of minimizing drag and maximizing maneuverability. Surrogate models for the two optimization objectives above were built using Multilayer Perceptron (MLP) neural networks and ensemble learning methods respectively, and their performances were compared with traditional surrogate models. The optimization problem was solved using the Non-dominated Sorting Genetic Algorithm II. Comparative analysis of the initial and optimized AUV shapes demonstrates significant improvements in hydrodynamic performance, confirming the feasibility and effectiveness of the proposed method.

The identified nonparametric model was used to tune the control parameters of the ship motion. Local Gaussian Process Regression (LGPR) method was utilized to model and predict the ship motion based on the collected ship motion data. The tuning of control parameters was converted to an optimization problem. The fitness function was designed to improve the control performances, and Genetic Algorithm (GA) was applied to tune the control parameters. The simulation results demonstrate that the proposed method has satisfactory control accuracy and dynamic performances in the path following and obstacle avoidance control.

The Boussinesq model is a kind of wave model widely used in near-shore engineering, and its computational accuracy mainly depends on the basic performance of the model, while the upper bound of nonlinear application of the model has always been the focus of attention. In recent years, the two-layer Boussinesq model has gained great progress regarding theoretical properties, numerical modeling and applications. However, the value of its nonlinear upper bound has not been given in any literature. So in this study, the stream-function of the two-layer Boussinesq model was solved using a combination of genetic algorithm and Newton's method to determine the upper bound value of the model, considering the highest spatial derivatives of order 3 and 5. In the same way, the stream-function solutions of the corresponding one-layer Boussinesq model were derived. The numerical results show that the nonlinear upper bounds of the two-layer Boussinesq model with the highest derivatives of order 3 and 5 are H/L = 0.137 and 0.138. Compared with the one-layer Boussinesq model, the two-layer model has a greater water depth of applicability regarding strong nonlinear characteristics. The combination of genetic algorithm and Newton's method proposed in this study can provide some references for solving the stream-function waves of the related Boussinesq models.