Microbially induced calcite precipitation (MICP) is effective to strengthen the calcareous sand foundation of reef fortifications via in-situ construction,dynamic response,and the dynamic characteristics and numerical model of MICP-treated calcareous sand are essential to the reef defense design and damage evaluation.The MICP cementation experiment is conducted on calcareous sand from the South China Sea.The strain rate effect of MICP-treated calcareous sand is evaluated through quasi-static uniaxial compression and dynamic mechanical testing of a split Hopkinson pressure bar (SHPB).A user subroutine VUMAT is developed based on the microplane model M7 for calibrating the material parameters.Compared with the calcareous sand penetration experiments,the numerical simulations demonstrate an increase in the penetration resistance of calcareous sand.The results reveal that the unconfined uniaxial compressive strength of MICP-treated calcareous sand specimen is 12.31MPa,and the dynamic increase factors are 1.117,1.485,and 1.828,respectively,at strain rates of 426s^-1,1150s^-1,and 1712s^-1 on the Hopkinson bar.Strain rate effect has a negative influence on the penetration depth of M7 constructive model by 4.2% DOP contribution.Compared with ordinary calcareous sand,the depth of penetration of MICP-treated calcareous sand is reduced by 40.11% on average,and the average target static resistance increases from 5.21MPa to 11.89MPa,indicating a significant increase in the resistance to penetration.These findings provide the experimental data and simulation model reference for damage analyses of reef fortifications subjected to high-impact loadings.

Aircraft aerodynamic configuration design has long adhered to an experience-driven design philosophy,where the aerodynamic performance of the baseline shape often determines and limits the performance improvements achievable through shape optimization.This traditional design approach is incapable of yielding disruptive aerodynamic configurations.In recent years,with the emergence of aerodynamic topology design concepts,aerodynamic topology optimization methods have gained attention and are expected to be applied to the design of novel aircraft aerodynamic configurations.Focusing on aerodynamic topology optimization methods,this paper reviews their development in terms of topology representation methods and optimization algorithms,and analyzes typical application cases of existing methods in the design of novel aircraft aerodynamic configurations.First,the historical development of aerodynamic topology optimization methods is outlined.Second,recent advances in the application of aerodynamic topology to novel aircraft configuration design are summarized.Subsequently,the challenges and difficulties faced by aerodynamic topology optimization methods are discussed,including difficulties in handling high-speed compressible turbulent flows and insufficient smoothness at fluid-solid interfaces in aircraft flow topology.Finally,future research directions in this field are prospected.

To enhance the performance and optimization efficiency of blended-wing-body underwater glider (BWBUG),this study proposes a dual-layer multi-objective optimization method for shape design of BWBUGs.A fully parameterized model of the shape of BWBUG is established,an analysis platform for shape performance is built,and a performance database and high-precision neural network mapping models are established.These mapping models are used to study the correlations among design parameters and performance metrics.Furthermore,a dual-layer constrained multi-objective optimization algorithm is constructed to optimize the BWBUG shape by taking the lift-to-drag ratio and internal capacity as objectives.The optimization process is conducted using the developed mapping models.The results demonstrate the lift-to-drag ratio and internal capacity of the optimized BWBUG are improved by 31.49% and 11.33%,respectively.Additionally,the weighted score analyses under varying weight scenarios further validate the effectiveness and robustness of the proposed optimization method.

Focusing on the speed control degradation of reentry gliding vehicles during terminal guidance phase due to model uncertainty such as aerodynamic parameter perturbations,a robust bias proportional navigation guidance law with speed constraint is proposed.This guidance law can satisfy the accuracy requirements for target interception and the terminal speed constraint of reentry gliding vehicle,and has robustness against uncertainties such as perturbations in aerodynamic parameters.An error dynamics model of flight speed and ideal speed curve is established based on the control strategy of additional angle of attack deceleration.By introducing the theory of fixed-time convergence error dynamics and considering the model uncertainty caused by the perturbations of aerodynamic parameters during the flight,a speed constraint guidance law based on the fixed-time convergence error dynamics is derived.It is also proven that the speed error dynamics is converged under the bounded model uncertainty.The performance of the proposed guidance law is validated through numerical simulation,and its robustness against initial parameter and aerodynamic parameter perturbations is also verified through Monte Carlo test.

As a new type of special vehicle,the wheeled-legged platform (WLP) extends the degrees of freedom at the wheel ends,and has the potential of traveling at high-speed in complex terrain.However,the vertical-longitudinal coupled dynamics of two-stage actuation unit and the unmeasurable states pose significant challenges to speed tracking and terrain-adaptive compliance control on uneven roads.To address these issues,this paper proposes an integrated adaptive compliance control method based on state estimation.Firstly,the forward and inverse kinematic models of the WLP are developed based on its movement configuration,and the reference dynamic models for vertical vibration damping and longitudinal compliance are constructed using the virtual model approach.Then an integrated adaptive compliance control framework in vertical-longitudinal directions is established.It incorporates longitudinal terminal sliding mode control,vertical optimal vibration damping control,and Kalman filtering to finally achieve the speed tracking and terrain adaptation under a reference posture.Tests under various conditions validate the effectiveness and stability of the newly proposed integrated compliance control method.For example,the peak vertical acceleration under the bumpy road condition is less than 1.1 m/s².

To address the issues of slow convergence and poor path planning associated with the Deep Q-Network (DQN) algorithm for mobile robot path planning in large-scale complex unknown environments,a path planning algorithm combining Ant Colony Optimization (ACO) and DQN,termed ACOG-DQN,is proposed.Initially,the pheromone mechanism of ACO is introduced to facilitate the selection of potential paths with the goal of reaching the destination,thereby reducing the number of ineffective environmental explorations and determining the optimal path.Concurrently,the previous path selection experiences are filtered using a threshold to form a sample set for training the Q-network,which is then utilized to determine the optimal path for the mobile robot in the current environment.Finally,a path selection mechanism is designed where the weight of the Q-network’s optimal path increases over time,using the optimal paths determined by ACO and the Q-network,as well as those determined by random exploration,as candidates.This mechanism selects the current action,aiming to achieve a path that is ultimately decided entirely by the Q-network.Simulation and physical experiments conducted in three different complex environments demonstrate that the proposed ACOG-DQN algorithm exhibits superior performance in terms of convergence speed,path quality,and algorithm stability compared to the DQN algorithm,thereby validating the effectiveness of the proposed algorithm.

A novel bipedal mobile robot used in the dangerous and complex environments is proposed based on a 4-UPU+2-P parallel serial hybrid mechanism.Unlike traditional continuous gait patterns of bipedal robots,the proposed robot switches between dynamic and static platforms to achieve alternating movement and turning with just six actuators,following a “top platform+bottom foot” continuous gait.The degrees of freedom of the robot are analyzed using screw theory,the inverse kinematics is established using the closed-loop vector method,and the singularities and workspace of the robot are analyzed.The performance of the robot is analyzed using the motion/force transmission indices,and the performance distribution maps within the workspace are generated.By employing a particle swarm optimization algorithm and taking the global transmission index as the optimization objective,the key structural parameters are optimized to improve the motion/force transmission performance of the mechanism.To improve the stability of robot during motion,the simulation analysis of the robot in stair-climbing environments is performed based on the zero-moment method.The results indicate that the designed dual-platform bipedal mobile robot with a reasonable structure demonstrates a certain degree of stability in stair terrains,and possesses potential applications in complex and hazardous environments such as battlefield reconnaissance and post-disaster rescue.

To improve the real-time and adaptive performance of intelligent automobile trajectory tracking,a variable universe fuzzy adaptiv model predictive control (VUFAMPC) method is proposed.The strong tracking square root cubature Kalman filter (ST-SRCKF) algorithm is used to estimate the lateral force of tire to obtain the real-time tire lateral stiffness values.And then,based on the traditional model predictive control method of trajectory tracking,a variable universe fuzzy control (VUFC) method is used to design a variable universe fuzzy model predictive controller (VUFMPC),and the estimated tire lateral stiffness is used as the parameter for VUFMPC to achieve the adaptive correction of controller parameters by combing it with ST-SRCKF algorithm,thereby obtaining VUFAMPC.Last,the comparative analysis and verification are made through hardware-in-the-loop experiment.The experimental result show that,compared to FMPC and AMPC,the overshoot of VUFAMPC is reduced by 11.1% and 18.8%,and the transition time is reduced by 46.6% and 67.9%.The trajectory tracking test shows that,compared to FMPC、VUFMPC and AMPC,the maximum tracking error of VUFAMPC is optimized by 73.9%,68.7% and 24.8%,and the average tracking error is optimized by 72.3%,56.1%,28%.The conclusion indicates that VUFAMPC has better real-time and adaptive performance,while effectively balancing the trajectory tracking accuracy and driving stability of intelligent automobile.

Power loss is one of the critical parameters for evaluating the performance of integrated hydro-mechanical transmission device (IHMTD) in tracked vehicle.The efficiency of IHMTD is comprehensively assessed by predicting its power losses under different operating conditions,thereby ensuring the good mobility of tracked vehicles.This paper proposes a method based on stacked ensemble learning for predicting the power losses of IHMTDs under various working conditions.The efficiency data from 75 tracked vehicles IHMTD under multiple working conditions is used for power loss prediction,and the Random Forest,LightGBM,AdaBoost,CatBoost and XgBoost algorithms are integrated together through stacking to effectively predict the power losses of IHMTD.Additionally,SHapley Additive exPlanations (SHAP) values are employed to analyze the impacts of various factors and models on power loss prediction.The experimental results show that the root mean square error (RMSE) of power loss prediction is 6.6,and the goodness of fit (R²) reaches 0.976 on the training set; while on the test set,these metrics are 8.92 and 0.961,respectively.Further analysis reveals that the input torque is the primary factor affecting power losses,and within the stacked ensemble framework,the Random Forest algorithm contributes the most to improve the prediction accuracy.

The series-parallel hybrid powertrain combines the characteristics of series and parallel hybrid configurations,and its engine and motor are connected through a power coupling mechanism.It can provide power to the wheels independently or jointly,resulting in excellent fuel economy.However,the mechanical connection between the engine and transmission system in series-parallel hybrid vehicle often leads to poor NVH (noise,vibration and harshness) performance.To address this issue,this paper proposes an active vibration control strategy.First,the configuration and various operating modes of series-parallel hybrid vehicle are introduced,a torsional dynamics model and an engine fluctuating torque model are established,and the relationship between system torsional vibration response frequency and engine speed is analyzed.Subsequently,an adaptive multi-channel active vibration control strategy based on the notch filtered-x least mean square (FxLMS) algorithm is proposed.The proposed strategy is to use the motors as actuators to achieve the compensation for the torsional vibration at target points by adaptively adjusting the weight matrix.In order to improve the control accuracy,the multiple secondary paths in the series-parallel system is identified offline using finite impulse response (FIR) filters.The simulated results demonstrate that the proposed active vibration control strategy shows excellent effectiveness and stability.The strategy is used to improve the ride comfort of vehicle,extend the service life of components,and enhance the NVH performance of the entire vehicle.

The negative obstacles pose a significant threat to the driving safety of unmanned ground vehicles (UGVs) when they operate in unstructured environments.However,in the existing research on negative obstacle detection,the camera-based detection methods are not well suited for the detection of negative obstacles in unstructured environments with poor lighting conditions and complex backgrounds; and LiDAR(light detection and ranging)-based detection methods are mostly designed based on the characteristics of mechanical LiDAR harnesses and have poor universality.Therefore,a highly compatible and scalable negative obstacle detection method is proposed,which can adapt to different types of LiDAR and achieve high-precision negative obstacle detection.The proposed method involves preprocessing the point cloud data,extracts the ground regions of interest,and performs point cloud pose correction.The adaptive resolution polar coordinate rasterization technology is used to enhance the spatial representation capability of point cloud data.A negative obstacle grid feature descriptor is designed,and the potential negative obstacle regions are extracted from multiple features such as the hollow characteristics,height differences,and minimum height of point clouds.A multi-frame fusion strategy is introduced,and the high-precision negative obstacle surface occupancy range is outputed through map reprojection and Bayesian rule-based probability updates.The experimental results show that the proposed method is applicable to LiDARs with different scanning modes,and can effectively identify the negative obstacle areas in complex unstructured environments.

Proportional servo valves are the core components in defense heavy-duty hydraulic drive systems such as missile launcher,tank and artillery.Due to the fact that the spool motion system is constrained by the complex dynamic characteristics such as hysteresis,flow force and Stribeck friction,the conventional proportional integral derivative (PID) control method is unable to overcome the inherent performance limitations.An adaptive active disturbance rejection controller (AADRC) for electro-hydraulic proportional servo valves is proposed,which integrates a parameter adaptation law with an extended state observer (ESO).A state-space model of the position loop for spool motion is established,revealing its nonlinear dynamic coupling mechanisms.To address the parameter uncertainties such as hydraulic valve spring stiffness and viscous damping coefficient,a parameter adaptation law is designed for online estimation of unknown parameters,and an ESO is utilized to compensate dynamically for the coupled nonlinear disturbances arising from hysteresis,flow force and unmodeled friction.Based on Lyapunov stability theory,it is rigorously proved that the controller is used to guarantee the global uniform ultimate boundedness (GUUB) of the closed-loop system’s states,parameter estimation errors and disturbance observation errors.Experimental results demonstrate that the proposed AADRC exhibits the smallest steady-state tracking error under both low-speed (0.5Hz) and high-speed (2Hz) sinusoidal conditions compared to PID controller and nominal model-based robust controller,and effectively overcomes the lag during reversing in triangular wave conditions.

To fundamentally resolve the fluid leakage in the traditional hydraulic artillery recoil mechanism,the magneto-mechanical stability of neodymium iron boron (NdFeB) high-strength magnetic materials under artillery impact environment is studied to achieve the innovative applications of eddy current principle in recoil systems.In view of the situation of unclear dynamic magneto-mechanical degradation law of NdFeB,a magneto-mechanical constitutive model suitable for brittle materials is established based on thermodynamic theory and damage evolution.The dynamic mechanics and impact demagnetization experiments are carried out,and the key parameters of the magneto-mechanical constitutive model are identified on the basis of the experimental data.A dynamic calculation model of artillery is built,and the proposed constitutive model is embedded in the dynamic explicit calculation via the subroutine.The magneto-mechanical degradation data of NdFeB at different positions and different times is extracted and analyzed.The results show that the most serious degradation occurs in the inner ring of the rightmost permanent magnet,and the maximum demagnetization is 0.1668T.This study not only gives effective guidance for the use of the electromagnetic brake,but also provides an important theoretical support for the application of NdFeB materials in other impact environments.

In order to meet the needs of modern urban combat and terminal air defense precision damage,and further improve the ranging accuracy of airburst munitions,the influence mechanism and factors affecting the ranging accuracy by counting the number of projectile revolutions are analyzed and studied using a modified prime ballistic equation.Three levels of nine factors:mass,initial velocity,lead,rotational inertia,temperature,air pressure,humidity,crosswind,and tailwind in the ballistic model are simulated through orthogonal test,revealing that the variations in lead caused by rifling manufacturing tolerances,rifling type,and wear are critical yet overlooked factors in existing ranging algorithms.An equivalent lead calculation method is proposed to address the impact of lead on ranging accuracy.This method defines the equivalent lead as the ratio of projectile velocity to rotational speed,which is calculated using an interior ballistic model.The effects of rifling type and wear are taken into account in the interior ballistic model.Simulations of the interior ballistic model demonstrate that the average difference between the equivalent lead and the theoretical lead is approximately 3%.Furthermore,the accuracy of the equivalent lead is validated through the simulations of the ranging algorithm,showing a significant improvement in accuracy.Specifically,at a fixed distance of 1000m,the use of equivalent lead to determine the number of loading revolutions results in an average increase in ranging accuracy of 32.9m.

It is required to fastly and accurately predict the exterior ballistics and impact points of mortars on the modern battlefield.A mortar exterior ballistics prediction method based on Transformer-long short-term memory (Transformer-LSTM) hybrid neural network is proposed.The Transformer network is utilized to extract the intrinsic joint features of mortar’s velocity and three-dimensional coordinates at the moments from T to T+K,and the LSTM network takes these time series features as an input to map the three-dimensional coordinate information at the moment of T+K+1.In order to optimize the network model,the effects of different sliding window step sizes on the convergence performance of the exterior ballistics prediction model are investigated and analyzed.The proposed hybrid network is compared with GRU and LSTM networks in terms of single-step,multi-step and impact-point prediction.It is found that the prediction accuracies of the proposed hybrid network for the three-dimensional coordinates of exterior ballistics can reach up to 99.78%,99.72% and 99.81%,which are better than those of GRU and LSTM networks; and the single-step prediction of the exterior ballistics of the proposed hybrid network consumes only 1.2ms,which significantly improves the prediction accuracy and efficiency.The method enables accurate and fast prediction of exterior ballistics and impact point,providing more response time for mortar interception missions.

Aiming at the issues of strong subjectivity in control model selection during the design of automatic chain rotational shell magazines,which lead to insufficient model accuracy and difficulties in improving control performance,a tripartite solution of “full-order modeling,parameter identification,and sensitivity-driven model reduction” is proposed.Based on the structural characteristics of a chain rotational shell magazine in an automatic loading mechanism for self-propelled artillery ammunition,a comprehensive nonlinear full-order dynamic model is developed by integrating critical factors including the dynamic characteristics of drive motors,transmission backlash of reducers,polygonal effect in chain drives,time-varying meshing collisions,nonlinear friction dissipation,and structural flexibility.To address the challenges of numerous coupled parameters and empirical calibration difficulties in the dynamic model,a multi-objective parameter identification framework based on the Particle Swarm Optimization (PSO) algorithm is developed for key parameter identification.Multi-condition experimental verification demonstrates that the steady-state relative error between the established full-order dynamic model and the actual system remains below 5.97%.Furthermore,Sobol global sensitivity analysis is employed to quantitatively evaluate the impact of 17 parameters on system dynamic response and positioning accuracy.The analysis reveals that the current-loop global sensitivity indices for load end mass and Coulomb friction are 0.25 and 0.42 respectively,while the position-loop global sensitivity indices for gear backlash and maximum static friction at the load end are 0.20 and 0.78 respectively.Based on these findings,a reduced-order mathematical model is constructed by retaining critical parameters,significantly simplifying controller design complexity and providing a lightweight model architecture with enhanced adaptability for model-based control algorithm development.

The influence of aluminum powder content on the energy output characteristics of RDX-based aluminized explosives is studied.The free-field static explosion tests are conducted on aluminized explosives with aluminum powder contents of 20%,30%,and 40%.The overpressure-time curves of free-field and ground shock waves are measured using overpressure sensors,and the propagation processes of incident waves,reflected waves,and Mach waves,as well as the changes in the explosion fireball are captured using a high-speed camera.The JWL-Miller parameters of the aluminized explosives are calibrated based on the test results.The test results show that the peak overpressure and specific impulse of ground reflected wave of the aluminized explosive are slightly affected when the aluminum powder content increases from 20% to 30%.When the aluminum powder content increases from 30% to 40%,the peak overpressure of ground reflected wave is decreased by 2.17% to 7.81%,and the specific impulse is decreased by 0.29% to 12.17%.When the aluminum powder content increases from 20% to 40%,the maximum diameter of explosion fireball increases from 6.48m to 7.12m,and the time when the explosion fireball begins to shrink increases from 20ms to 60ms.The maximum errors of the blast shock wave parameters (peak overpressure and specific impulse) of 30% and 40% aluminized explosives predicted using the JWL-Miller parameters are 16.75% and -8.51%,respectively,compared with the test results,and the average absolute percentage errors are 6.76% and 4.14%,respectively.This study can provide reference and guidance for the prediction of the power field of aluminized explosives.

The micro-scale transient combustion characteristics and flame structure of nitrate ester plasticized polyether (NEPE) propellant under rapid depressurization are investigated.In this study,the heterogeneity of NEPE propellants is characterized by introducing the randomly distributed particle filling structures,and a semi-global gas-phase chemical reaction kinetic model is constructed to simulate the combustion process of NEPE propellants.The results show that the flame structure changes continuously in the initial stage of combustion,showing a significant oscillation characteristic lasting for about 3.5ms.The oscillation is due to the interaction between particle distribution and depressurization.When the decompression process lasts for 19ms,the pressure drops to about 0.1MPa,and the flame exhibits a premixed structure.The heterogeneity and pressure change of NEPE propellant lead to the unstable flow of gas-phase components on its surface,thus forming a dynamic flame structure.

Ballistic helmets are critical protective equipment designed to mitigate craniocerebral injuries caused by ballistic impacts and blast shocks.The cushioning system within the helmet plays a vital role in attenuating the damage effects from such impacts.To enhance the overall protective performance of headgear,a novel cushioning material—EVA/SSG composite foam—was developed by incorporating Shear Stiffening Gel (SSG) into ethylene-vinyl acetate (EVA) foam.Using a ballistic impact testing platform and a synthetic head model,comparative ballistic tests were conducted on helmets equipped with conventional EVA pads and those with the EVA/SSG pads.The temporal variations in impact force on the headform surface and the acceleration at the centroid of the head model during impact were investigated.Additionally,finite element simulations of the ballistic impact were performed.Both experimental and numerical results demonstrate that,under ballistic impact conditions,the helmet with EVA/SSG foam liners reduces the peak pressure on the headform surface and the peak acceleration at the centroid by more than 20%,thereby significantly lowering the probability of head injury.

In order to investigate the performance of different clustering algorithms in processing the finite element resultant stress nephograms,a single-layer ballistic impact finite element model of Twaron^® plain weave fabric is established.Four different clustering algorithms,namely,k-means,Gaussian mixture model (GMM),Mean-shift,and density-based spatial clustering of applications with noise (DBSCAN),are used to cluster the stress nephograms and analyze the results comparatively by taking the resultant images of the stress distributions as an example.The results show that the Mean-shift and DBSCAN clustering algorithms are not suitable for processing a large number of finite element stress stress mapss,the k-means and GMM clustering algorithms improve the processing efficiency by 74.24 and 172.64 times compared with the traditional manual processing,and the GMM clustering algorithm produces errors when the color of the image is not clearly differentiated.The k-means clustering algorithm ensures high efficiency while keeping the error within 0.85%.Therefore,among these four algorithms,the k-means clustering algorithm is the most suitable for fast,objective and quantitative analysis of a large number of stress nephograms.Using k-means clustering algorithm,it is measured that the area of the stress interval of single-layer Twaron^® plain weave fabrics decreases with a stress area change rate of 5.61×10⁷mm²/s for 0-600MPa,and the area of the stress interval increases with a stress area change rate of 5.27×10⁷mm²/s for 600~1200MPa within 1-15μs of the impacts.

This paper proposes an adaptive key edge detection framework based on deep reinforcement learning to address the challenge of identifying the key edges in space-based kill chain network.This method first uses complex networks to model a space-based kill chain.And then a template-based kill chain search method is proposed by introducing the idea of backtracking search.A key edge exploration paradigm is constructed based on deep reinforcement learning,which combines a layered experience replay mechanism with a dynamic ε-greedy strategy.,The precise localization of key edges is achieved through multi-dimensional state representation.The experiment shows that,in the simulation of space-based kill chain network testing,the Top-10 recognition accuracy of the proposed method reaches 85%,which is 89.5% higher than that of the traditional betweenness centrality method and 19.7% higher than the benchmark of deep Q-network (DQN).In terms of network robustness indicators,the global efficiency reduction rate η is 37.4% higher than the edge betweenness centrality,and the prediction error of the maximum connected component retention rate is controlled within 5%.

The dynamic mechanical property of fiber-reinforced polyurea material is studied.The dynamic compression experiments are conducted on the polyurea material specimen by using a separated Hopkinson pressure bar (SHPB).The stress-strain curves of the specimens within different strain rates at three different temperatures are obtained through the experiments.A visco-elastic constitutive model based on the standard linear solid model is established.The experimental results show that the dynamic mechanical response of fiber-reinforced polyurea is nonlinear,with the stress-strain curve consisting of a super-stress region and a hyperelastic region.As the strain increases,the super-stress value tends to stabilize,and the nonlinear term associated with strain rate in the constitutive model also gradually approaches a stable value related to the strain rate.The visco-elastic constitutive model based on the standard linear solid model can accurately describe the dynamic mechanical behavior of fiber-reinforced polyurea.The constitutive model is implemented through secondary development in LS-DYNA for simulation.In both SHPB and air blast simulations,the simulated results are in good agreement with the experimental data,which validates the accuracy of the constitutive model.

In order to characterize the mechanical response of Q345E steel under high-speed impact conditions,the quasi-static tensile and Hopkinson compression experiments are conducted to determine the static and dynamic mechanical properties of Q345E steel as well as its fracture characteristics in wide stress triaxiality.Based on the Johnson-Cook (JC) and Cowper-Symonds (CS) models,a JC-CS constitutive model is proposed to characterize the static and dynamic mechanical behaviors of Q345E steel.A three-stage Wierzbicki fracture model is peoposed and the failure parameters in the range of each stress triaxiality are obtained through experiment and numerical simulation.The Abaqus VUMAT subroutine is used to verify the validity of the constitutive and fracture models.The results show that Q345E steel has obvious strain hardening and strain rate effects,and the two kinds of effects are coupled under dynamic loading.The proposed JC-CS constitutive model can accurately characterize the rate-dependent strain hardening and strain rate effects in the range from quasi-static to high strain rate.The fracture strain of Q345E steel is related to the stress triaxiality.When the stress triaxiality is greater than -1/3 and less than 2,the fracture strain decreases first and then increases,and finally tends to zero.The fracture strain becomes infinitely large when the stress triaxiality tends to -1/3.The numerically simulated results demonstrate that the fracture morphology,the failure mode of target plates and the process of penetrating a multi-layer target predicted by the simulation are highly consistent with the experimental results.This indicates that both the proposed JC-CS constitutive model and Wierzbicki fracture model effectively capture the mechanical properties and fracture behavior of Q345E steel as well as its response characteristics under high-speed impact.

To address the challenges of insufficient data,complex feature selection,and limited applicability of traditional empirical formulas in the prediction of concrete penetration depth,this paper proposes a concrete penetration depth prediction model based on deep learning and optimization strategies.Initially,218 sets of concrete penetration test data are collected,and the size and diversity of dataset are expanded through the integration of multi-stage empirical formulas.To enhance the accuracy of feature selection,an XGBoost (eXtreme Gradient Boosting) model combined with SHAP (SHapley Additive exPlanations) values is employed for feature screening and dimensionality reduction,thereby identifying the key features that are correlated with penetration depth.A multi-layer perceptron (MLP) model for penetration depth prediction is constructed based on these selected key features.Further,Bayesian optimization method is applied to fine-tune the hyperparameters of the MLP model,and its generalization capability is evaluated through K-fold cross-validation.The results demonstrate that the proposed model performs better than the traditional empirical formula under various conditions,its prediction accuracy and stability are significantly improved,and it shows good generalization ability and adaptability.In addition,compared with the traditional empirical formulas,the proposed model also shows better prediction effect in the applicable range of the empirical formulas.

The underwater unmanned vehicle (UUV),as an intelligent underwater operational system,generates a noise during operation,that can significantly interfere with the flank array sonar.The current mainstream adaptive beamforming techniques are primarily designed to suppress the far-field interference and have insufficient adaptability to the strong near-field interference caused by UUV self-interference.To address this issue,this paper proposes a UUV self-interference suppression algorithm based on covariance matrix reconstruction.The covariance matrix of the UUV self-interference exhibits high temporal stability under constant operating conditions due to the rigid connection between the flank array and the UUV.Leveraging this characteristic,the peoposed algorithm first estimates the self-interference covariance matrix with pre-collected data,and then utilizes this matrix to modify the covariance matrix of far-field interference and noise.Consequently,an interference-plus-noise covariance matrix is constructed to compute the weights of the adaptive beamformer.Experimental results indicate that,compared to current mainstream adaptive beamforming methods,the proposed algorithm performs better when the input signal-to-noise ratio exceeds 0dB.Moreover,compared to matrix filtering methods,it can enhance interference suppression effect by approximately 6dB.

The proposed method addresses the issue of insufficient conceal and poor robustness in path planning for underwater vehicles in complex marine environments,where current methods only consider hydrological conditions for stealth planning.This method first calculates the instantaneous detection probability of the underwater vehicle influenced by the hydrological environment by integrating the sonar equation with the Bellhop model and using temperature and salinity data.It then incorporates the models of target lateral distance and patrol width,combined with terrain data,to obtain the probability of detection platforms constrained by geographical factors and the sonar contact time.Finally,it integrates these two aspects to construct a concealment effectiveness model that takes into account the influence of multiple environmental factors.Finally,an A^* algorithm that considers concealment is used to automatically calculate the concealed route.Experimental results show that,compared to traditional concealment planning methods,the proposed method integrates hydrological and geographical factors into a normalized stealth effectiveness value and applies it to the stealth planning algorithm,which significantly reduces the average cumulative detection probability of the vehicle when navigating complex straits,enhancing the concealment of the planned route; results in a more robust concealment route that better meets practical application requirements.

The crew is the weakness of the ship anti-impact system, and it is helpful to improve the protection design by clearly defining the impact response characteristics of human body.This paper focuses on the impact response regularities of crew in different postures.The impact response characteristics of supine human are studied via the real ship-human underwater explosion (UNDEX) impact test for the first time; A far field UNDEX load-structure-human body integrated response model is constructed based on the Taylor fluid-structure coupling principle, and the vertical response differences of typical organs of human body in different postures under the conditions of typical impact factors (0.30, 0.45 and 0.60) are compared and analyzed.The results indicate that the vulnerable organs of sitting, standing and lying persons are the pelvis, feet and head, respectively, which should be given priority protection.When the human body changes from a sitting position to a standing position, the loads on the organs in the upper part of human body significantly decrease, and the lumbar spine and pelvis are no longer vulnerable parts prone to injury.There is a tendency for the vulnerable parts to shift from the pelvis to the head when the human body changes from a sitting position to a lying position.Under the same impact factor, the response degrees of the organs of human body in different postures directly contacted with the deck are more sensitive to the change of deck mass and show a negative correlation.The research results can provide technical support for the impact protection design.

The operational safety and catalytic efficiency of hail-suppression and rain-enhancement shells,as critical tools for weather modification,are directly affected by the fragmentation characteristics formed after self-destruction.This paper investigates the mechanism of shell material mechanical properties effecting on fragment formation through experiments,theories and numerical simulations.The differences in fragment mass distribution,morphological characteristics and energy dissipation among S20 steel,9260 steel,D60 steel,and 823 steel are analyzed.The results reveal that the strength-toughness synergy of shell materials plays a decisive role in fracture patterns.High-strength materials (e.g.,9260 steel) generate uniform fine fragments (>60% of 1-5g fragments) but exhibit localized energy concentration due to insufficient toughness.Low-strength high-toughness materials (e.g.,D60 steel) produce irregular agglomerated fragments (32% of more than 10g fragments) due to significant plastic deformation.In contrast,823 steel demonstrates optimal brittle-ductile fracture coordination under explosive loading due to its unique mechanical properties (1000-1200MPa tensile strength and 40-60 J impact toughness),with 85% of less than 5g fragments and over 90% of ≤10g fine fragments,which fully complies with national safety standards (Class B,≤10g) and significantly reduces ground safety risks.The research provides theoretical guidance for optimizing the hail-suppression and rain-enhancement shell materials and holds substantial engineering value for enhancing the weather modification safety.

The high resolution range profile (HRRP) is predominantly applied in the field of radar target recognition because of providing the detailed information of target features.The traditional radar HRRP recognition method of ballistic midcourse targets is affected by the environmental noise interference,atmospheric radiation and penetration strategies,resulting in low target recognition accuracy.Furthermore,the intelligent optimization algorithm faces the challenge of an extensive number of parameters when extracting the local features of a target,which makes it difficult to perform manual parameter adjustment.Aiming at this issue,a high-resolution range profile recognition method based on the improved grey wolf optimizer and one-dimensional convolutional neural network (IGWO-1DCNN) is proposed for ballistic targets.An improved 1D convolutional neural network is constructed for the feature extraction from HRRP samples of wideband radar targets.The improved grey wolf optimizer (IGWO) algorithm is introduced to accelerate the convergence speed and recognition performance of the model.The support vector machine (SVM) is used as the classifier to facilitate the recognition processes.The experimental results demonstrate that the proposed method is capable of accurately identifying the ballistic targets,automatically optimizing the parameters of neural network,and reducing the burden of manual training and exhibiting higher robustness.

Infrared small target detection has extensive applications in military fields such as infrared guidance and tracking systems and is an important area of infrared image processing.Due to the limitations in detection equipment and the lack of inherent information about infrared small targets,the existing detection methods are difficult to meet the practical performance requirements.In order to explore a lightweight and highly accurate infrared small target detection model,a lightweight and efficient YOLOv10n infrared small target detection model (L-YOLOv10n) is designed based on YOLOv10n.The SCDown module in YOLOv10n is replaced by a lightweight spatial-channel decoupled downsampling (L-SCDown) module to enhance the key features of infrared small targets with a low computational cost.A lightweight Cross-stage partial convolution with Two Fusion layers (L-C2f) module is used to replace the C2f module,thereby enhancing the edge information of small targets and extracting the multi-scale features while reducing computational cost.To address the issues of infrared small targets with few pixels and an imbalance between foreground and background,Focal Loss and a focaler intersection-over-union (Focaler-IOU) loss function are introduced,thus allowing the model to better focus on the difficult-to-detect targets.Experimental results on the public datasets SIRST-V2 and NUDT-SIRST demonstrate that L-YOLOv10n significantly outperforms the detection-based models in both detection performance and resource consumption.The detection performance of L-YOLOv10n is slightly lower than that of Transformer-based segmentation models,but its resource consumption is significantly better than those of other models.Its generalization performance on the NUDT-SIRST dataset is also significantly higher than those of most infrared small target detection models.These results demonstrate that the proposed model strikes a balance between resource consumption and high-precision detection,demonstrating its practicality.

Transformer is a kind of mainstream method used for infrared and visible image fusion (IVIF).However,most of the Transformer-based methods for IVIF suffer from the problems such as large number of parameters and high computational complexity.To this end,a method for infrared and visible image fusion based on Mamba-empowered triple-branch generative adversarial network is proposed.Specifically,three independent and cooperative branches for feature extraction and fusion are designed in the generator network.The infrared and visible branches are utilized to extract the global features from the infrared and visible images,respectively,through the shallow feature extraction modules and Mamba blocks.Meanwhile,the global features extracted by the infrared and visible branches are hierarchically integrated into the convolutional neural network-based intermediate fusion branch,thereby achieving the full interaction and fusion between the local and global features.Furthermore,the generator is competitively trained against two discriminators (infrared discriminator and visible discriminator),forcing the generator to improve its ability to produce fusion image.The experiments on public datasets indicate that the proposed method outperforms other methods both in the qualitative visual effects and the quantitative objective metrics.