The control strategy of traditional tank bidirectional stabilizers is difficult to effectively deal with the coupling,nonlinearity and uncertainty in the new generation of all-electric bidirectional stabilizers,while the model-based nonlinear control can make full use of a priori information from the dynamic model of the system to enhance the control effect.Based on this,an electromechanical coupled dynamics model of an all-electric bidirectional stabilizer taking the actuator dynamics into account is established,and a nonlinear sliding mode control method based on neural network compensation is proposed.The sliding mode surface and the improved sliding mode robust control law based on hyperbolic tangent function are introduced to construct the nonlinear sliding mode control function,which can effectively eliminate the system oscillations and improve the system stability.Meanwhile,the multilayer neural network technique is deeply fused to accurately estimate the uncertainty in the system and make compensation for the feedforward,thereby avoiding high-gain feedback.It is rigorously demonstrated by the stability theory based on Lyapunov functions that the proposed control strategy can achieve the asymptotic stability performance of tank all-electric bidirectional stabilizer with continuous control inputs.A co-simulation environment and a semi-physical experimental platform are built.The superiority of the proposed control strategy is verified through a large number of comparative experiments.

The accurate delay performance of digital electronic detonator provides the conditions for selecting a method of hole-by-hole detonation in small section roadway,but the selection of delay time between holes has always been a difficulty.Based on the theory of rock breaking and millisecond blasting,an accurate delay calculation method of hole-by-hole detonation in roadway is constructed.The three-dimensional numerical model of full section is established by using the generalized discrete element method (GDEM) software.Based on the calculated delay time,the hole-by-hole detonation is numerically simulated to analyze the damage and effective stress of roadway section.The results show that,within 50ms,the cutting holes are all detonated,the delay time between the holes is short,the superposition of stress waves is obvious,and an obvious crushing zone is generated.After 50ms,the auxiliary holes and the surrounding holes are detonated one by one,the free surface is continuously expanded,the clamping force is reduced,the delay time between the holes is increased,the superposition of stress waves is weakened,the rock is broken evenly,and the explosion energy is more fully utilized.The hole-by-hole blasting test results of underground mine roadway are remarkable,there are no obvious chunks after blasting,the average block size is 19 cm,and the average hole utilization rate is 93.3%.The precise delay hole-by-hole blasting method effectively improves the tunneling efficiency of small-section roadway and controls the blasting disturbance.

This paper investigates the trajectory planning issue for fixed-wing UAVs and proposes a real-time trajectory planning method based on exact convex relaxation.This method includes two steps,i.e.path planning and velocity optimization.In the first step,a flight path of UAV in the multi-obstacle environments is designed.In consideration of the dynamics and control constraints,an obstacle avoidance path planning method based on Dubins path is proposed to generate a flyable obstacle avoidance trajectory.In the second step,the velocity and control of a fixed-wing UAV flying along an obstacle avoidance path are calculated such that it can track the obstacle avoidance path.The nonlinearity retention and exact convex relaxation are used to convert the highly nonlinear velocity optimization problem into one single convex optimization problem,and the validity of the proposed method is theoretically proven.Since there is no iterative process of solving the convex optimization problems,the velocity optimization algorithm does not have convergence issues and has remarkable real-time performance.The simulated results demonstrate that the proposed method can realize obstacle avoidance trajectory planning reliably and rapidly in both multi-obstacle environment and unknown obstacle environment,and it significantly improves the computing efficiency compared with the nonlinear programming and successive convex optimization methods.The calculation time is only tens of milliseconds.

The current radar jamming decision-making method based on reinforcement learning sets the exploration rate parameter according to a single factor and fixed law,which leads to the increase in the number of confrontation rounds required for algorithm convergence.A reinforcement learning-based radar jamming decision-making method with adaptive setting of exploration rate is proposed.Based on the Metropolis parameter adjustment criterion of simulated annealing method,an adaptive setting criterion of exploration rate is derived from the number of radar operating states recognized by jammers,the number of jamming successes,the change rate of algorithm convergence curve and the jammer’s cognition of radar in the process of countermeasures.According to the effectiveness of jamming action,a jamming action space clipping strategy is designed to reduce the dimension of jamming action space and further improve the convergence speed of the algorithm.In the simulation experiment,two different radar working state diagrams are designed and compared by using the Q-learning algorithm.The simulated results show that the proposed method can achieve the adaptive setting of exploration rate when the radar working state transition relationship changes.Compared with the exploration rate setting scheme based on simulated annealing method,single factor and fixed law,the number of confrontation rounds required for the convergence of the proposed method in the two state diagrams is reduced by 18%,26%,45% and 42%,44%,48%,respectively.At the same time,it can also obtain greater benefits and higher jamming success rate,which provides a new idea of exploration rate setting for multi-functional radar jamming decision-making based on reinforcement learning.

The prismatic charge structure,as a typical asymmetric structure,exhibits a pronounced directional energy output.It is important to predict the velocity distribution of its fragments for the structural design and damage efficiency assessment of warhead.For the prismatic charge structure,a fragment velocity prediction model based on an artificial neural network (ANN) is proposed.To improve the predictive efficiency and accuracy of the network model,the key factors affecting the fragment velocity distribution are identified through theoretical analysis,and four input characteristic parameters are selected for the network model.The multiple sets of different numerical simulation conditions are established by adjusting the values of these characteristic parameters,and a dataset is provided for the network model through numerical simulation method.The trained network model is used to predict the test set,and the predicted results are in good agreement with the numerically simulated results.The results indicate that the proposed network model has high accuracy in predicting the fragment distribution of prismatic charge structure,and has a good generalization capability.The neural network model is characterized by fast computation speed,high predictive accuracy,and easy-to-modeling.It can accurately predict the fragment velocity distribution of prismatic structure under single-end initiation conditions,thus providing important data support for the structural design and damage efficiency assessment of warheads.

Metal halide perovskites have become a current research hotspot owing to their exceptional optoelectronic properties and significant potential application value. The elemental composition and crystal structure of perovskite material have crucial influence on its performance.To realize the preparation of novel perovskites,this paper focuses on the research of synthesizing cesium lead chloride perovskite (CsPbCl₃) under shock loading.In this study,CsPbCl₃ perovskite powder is synthesized by the shock loading of detonation-driven flyer plates under the conditions of 0.6-0.8 relative densities of powder and 14.2-27.9GPa shock pressures.The characterization results of X-ray diffraction (XRD),scanning electron microscopy (SEM) and transmission electron microscopy (TEM) indicate that the recovered products are CsPbCl₃ perovskite powders.Experimental results also demonstrate that the shock pressure and the relative density of precursor are two key factors to the synthesis of CsPbCl₃ perovskite powder.Based on the experimental conditions and analytical characterization results,a formation mechanism of CsPbCl₃ synthesis is discussed.It is confirmed that the proper shock pressure for high-purity CsPbCl₃ powder synthesis is from 14GPa to 17GPa.It is also suggested that the shock synthesis method is a feasible approach for preparing difficult-to-synthesize perovskite.

With the rapid development of unmanned aerial vehicle (UAV)-related technologies,the intelligent UAV swarms have attracted more and more research attention.The organizational algorithm models of natural swarms are used for the intelligent UAV swarms,and combined with other intelligent technologies to form advanced group intelligence behaviors,which have the advantages that traditional single artificial intelligent agents cannot match.However,the issue of reliability is also an important factor that restricts its large-scale application.The concept,characteristics,and technological development path of intelligent UAV swarms are systematically analyzed.Based on the application scenarios,application modes,and behavioral processes of intelligent UAV swarms,the connotation of UAV swarm reliability is discussed.A technological framework for the reliability of UAV swarms is constructed,and the technical approaches and suggestions for the development of UAV swarm reliability technology are given based on the technological characteristics and application requirements of intelligent UAV swarms.Under the reliability connotation of “one core, two capabilities, and N foundations,” a reliability technology framework for intelligent UAVs has been proposed,which consists of reliability metrics,reliability development and verification,and reliability maintenance and support technologies.

In order to study the effect of horizontal inflow on the aerodynamic characteristics of small unmanned aerial vehicle (UAV) rotors,the wind tunnel test and numerical simulation are conducted to obtain the multi-dimensional aerodynamic data and wake vortex diffusion characteristics of UAV rotors under different inflow speeds.A low-speed wind tunnel is used to simulate horizontal inflow,and a six-component force-moment sensor is used to measure the aerodynamic forces the rotational speeds from 2000r/min to 9000r/min.The experimental study focuses on analyzing the variations of lateral force,and pitch and roll moments in the rotational plane.The influence of inflow speed on surface pressure distribution during the rotational period and the morphological characteristics of the near-field wake vortex are discussed from the numerically simulated results.The results indicate that the optimal rotational speed range of rotor is 4000-6000r/min under the action of horizontal inflow The x-axis force in the rotational plane is the primary contributor to the lateral force,with both x-axis and y-axis moments playing roles simultaneously.The lateral forces and moments become significant at inflow speeds above 5m/s and must be considered in dynamic analysis.The horizontal inflow disrupts the symmetrical feature of the rotor wake vortex.The inflow with greater speed leads to a larger inclination angle,causing the diffusion evolution of wake vortex to be compressed.

The reactive penetrator with enhanced lateral effect (PELE) exhibits a violent deflagration reaction when penetrating the reinforced concrete target,and the opening effect of it penetrating into the reinforced concrete target is more obvious than that of the inert filling PELE.To reveal the enhanced opening mechanism of reactive PELE,an enhanced opening model of reactive PELE penetrating into the reinforced concrete target is presented,which fully considers the radial rarefaction and the deflagration reaction of filling.The experiments of reactive PELE penetrating into the reinforced concrete targets at five impact velocities are carried out.The reaction characteristics of filling,the deformation angle-length evolution of shell,and the enhanced opening behavior are discussed based on the experimental and theoretical results.The results show that the proposed model can effectively predict the shell deformation and the target opening diameter,which the average errors are 6.9% and 8.5%,respectively.The reactive PELE shell exhibits double bending deformation and curling deformation modes,and the deformation angle-length evolution of shell is given.When the impact pressure exceeds than 2.32GPa,the neglect of the radial rarefaction wave may overestimate the reaction level of reactive PELE filling and the radial deformation of shell.The deflagration reaction of the reactive PELE filling increases the maximum opening diameter by 24.3%.

To reveal the mechanism of deflagration formation in high-energy propellant charge subjected to fragment impact,the semi-perforation and penetration of fragments into high-energy propellant charges are tested.The response types of charge are combustion-to-deflagration transition,combustion and deflagration.Based on the time-series image of charge reaction evolution and recovered samples,the response process of charge is comprehensively characterized using the image digital processing technology.The mechanical deformation and charge response characteristics of propellant after penetration are analyzed,and the deflagration mechanism of charge during semi-perforation and penetrating is revealed.The results indicate that,the trajectory penetration damage of the high-energy propellant charge impacted by a tungsten alloy fragment with a diameter of 10 mm is local radial cracking,and is accompanied by a delayed deformation caused by viscoelastic dissipation.There is a delayed reaction of the charge during semi-perforation and penetrating.that is,the significant combustion reaction of the charge occurs after the failure of propellant structure.During semi-perforation,the delayed deformation causes the high-intensity reaction zone inside the charge to move forward and coupled with structural constraints,resulting in combustion to deflagration transition.The charge is difficult to undergo deflagration reaction when delayed deformation and delayed reaction are significantly present in the penetrating state.When the combustion gas inside the crater seeps into the unreacted condensed phase propellant through the radial cracks of penetration trajectory,the charge usually undergoes a deflagration reaction.

In response to the challenge of inadequate target recognition capabilities due to the limited detection dimension and weak azimuth resolution of conventional frequency modulated continuous wave (FMCW) fuze,a fuze target recognition method based on motion array microwave imaging and multi-scale deformable convolutional networks (MSDCN) is proposed.A FMCW motion array antenna model is established thorough analysis of the thorough analysis of echo phase variation during the fuze motion.The virtual array elements of fuze antenna are expanded by motion synthesis to significantly enhance the azimuth resolution of the fuze,thus achieving the two-dimensional high-resolution imaging of target distance and azimuth.Simultaneously,a MSDCN target recognition model is constructed by delving into the multi-scale characteristics of the images formed due to the variations in target position,attitude,distance,and other states during the fuze-target encounter process.This enhances the adaptive recognition capability of fuze for the multi-scale characteristics of target imaging in complex dynamic encounter scenarios.The experimental results demonstrate that the proposed method significantly enhances the azimuth resolution of fuze.It achieves satisfactory imaging and recognition results in various target scenarios.The accuracy of multi-scale image recognition for typical targets reaches 94%,and even at -6dB signal-to-noise ratio,the target recognition accuracy remains at 88%.

A multi-stage decision problem of the dynamic weapon-target assignment (DWTA)is modelled and solved based on a game idea.For determining the targets to be attacked in dynamic multi-stage,the three-way decision theory is introduced to improve the threat assessment model,and the priority targets to be attacked in each stage are selected for firepower assignment through the acceptance,delay and reject decisions.A dynamic weapon-target assignment model based on the game idea is constructed,and the objective function of enemy avoidance risk value is designed to take into account the intelligence and uncertainty of the incoming target.For the influence of decision preference in multi-stage solution,a dynamic selection mechanism combining the crowding degree and reference point is designed to cope with the preference of different objective functions in dynamic situations and improve the solution quality.The improved algorithm is used to solve the proposed DWTA problem,and the effectiveness of the proposed algorithm is verified through comparative simulation experiments.

The two-mirror optical system is widely used in the fields and space remote sensing,detection,guidance and so on.Assembly is a key step that impacts the imaging quality of optical system.Currently,there is a lack of systematic research on the correlation between various assembly errors and imaging quality of optical system,which hinders real-time adjustment of optical system.In this study,a joint simulation method is proposed for the assembly and imaging of two-mirror optical system.This method involves using finite element simulation to obtain mirror surface figure errors,followed by precise fitting using Zernike polynomials.Subsequently,optical design and analysis software is employed to conduct the light path imaging simulations for the mirror surface figure errors fitted by Zernike polynomials and the assembly position deviations.The energy concentration degree serves as a quantitative evaluation index for assessing the imaging quality of optical system under different assembling error conditions.Furthermore,a support vector regression (SVR) surrogate model that incorporates both local and global mixed kernel functions is established to accurately capture the correlation between assembly errors and imaging quality.Research results indicate that the mixed kernel function SVR surrogate model proposed in the paper exhibits the smallest imaging quality prediction error with an average prediction error of only 6.51% compared to single kernel function or no kernel function SVR models.The proposed joint simulation method for assembly and imaging,along with the mixed kernel function SVR surrogate model,provides auxiliary support for real-time adjustment of optical systems under varying assembly error conditions.

The low-accuracy position estimation,large current ripples and weak stability in sensorless control of permanent magnet synchronous motors(PMSMs)based on lower-bridge current sampling topology under low-carrier ratio operations are resulted from the digital delay.For The issues above,a half-period calculation single-sampling double-update space vector pulse-width modulation(HSSDU-SVPWM) method is proposed.In this method, the phase current is sampled at a zero-crossing point of triangular carrier,and the target voltage vector is calculated in the half period after sampling.In considering the motor spin angle during digital delay,the target voltage vectors are outputted in the first half and the second half of the next pulse-width modulation(PWM)period,and are compensated with the equivalent delays of 0.75T_s and 1.25T_s,respectively,to reduce the error between the target voltage vector and the actual output voltage.Finally,the proposed method is applied to a sensorless control of PMSMs based on a model reference adaptive system(MRAS)speed observer.Comprehensive experiments verify that the proposed method can not only achieve more accurate position estimation and smaller current ripples,but also run stably at a rated load with a carrier ratio of 4.0.

Chaff centroid jamming of unmanned platform is an important means of missile terminal defense.The intelligent decision-making ability in platform maneuvering and chaff launching is an important factor to determine whether the strategic assets can be protected successfully.The current decision-making methods,such as computational analysis based on mechanism model and space exploration based on heuristic algorithm,have the problems of low degree of intelligence,poor adaptability and slow decision-making speed.A dynamic decision-making method of chaff jamming for terminal defense based on multi-agent deep reinforcement learning is proposed.The problem of cooperative chaff jamming of multi-platform for terminal defense is defined,and a simulation environment is constructed.The missile guidance and fuze model,unmanned jamming platform maneuvering model,chaff diffusion model and centroid jamming model are established.The centroid jamming decision problem is transformed into a Markov decision problem,a decision-making agent is constructed,the state and action spaces are defined,and a reward function is set.The decision-making agent is trained by using the multi-agent proximal policy optimization (MAPPO) algorithm.The simulated results show that the proposed method reduces the training time by 85.5% and increases the success rate of asset protection by 3.84 compared with the multi-agent deep deterministic policy gradient (MADDPG) algorithm.Compared with the GA,it reduces the deciding time by 99.96 % and increases the success rate of asset protection 1.12.

In order to quantitatively analyze the effect of explosion impact on the ignition energy of electronic detonator,seven sets of impact experiments with different strengths are made for liquid aluminum electrolytic capacitors by underwater explosion method.The dielectric breakdown behavior and leakage current change rule of capacitor under impact load are studied,and the energy loss path of electronic detonator capacitor is analyzed.An impact-ignition model of electronic detonator is established.The functional relationship between shock wave overpressure and ignition energy is obtained.The results show that the dielectric breakdown of capacitor sample occurs under the shock wave overpressure of 8.5-40.6MPa,and the voltage drop increases exponentially with the increase in shock wave overpressure.When the shock wave overpressure is greater than a critical value,the capacitor cannot heal completely after dielectric breakdown,and the leakage current increases to the mA level with an average value of 1.62mA.The ignition energy of electronic detonator decreases gradually with the increase in shock wave intensity,and when the leakage current increases,the ignition energy drops sharply.

The simultaneous localization and mapping (SLAM) algorithm has low positioning accuracy and cannot generate the dense maps in dynamic environment.For the above problems,a visual SLAM algorithm based on dynamic region exclusion and dense mapping is proposed.The algorithm creates a dynamic feature point detection thread into the original ORB-SLAM3 algorithm,and the YOLOX network is used to obtain the semantic information and object detection boxes in dynamic scenes.The algorithm detects the motion state of feature points by combining semantic and geometric constraints.A dynamic feature exclusion algorithm is proposed to accurately remove the dynamic feature points.Subsequently,a dense mapping thread is designed to construct dense point cloud maps based on keyframes and their corresponding poses.The remaining static feature points in the map are used to remove the ghosting caused by dynamic objects,thus achieving the construction of a dense map.The proposed algorithm is verified in the public TUM dataset and real dynamic environment.In the dynamic environment of TUM dataset,the proposed algorithm effectively eliminates the impact of dynamic objects on pose estimation,and improves the positioning and mapping accuracies of SLAM algorithm in dynamic scene.

In order to optimize the distribution of explosive energy in the smooth blasting of holes around underground projects, improve the hole mark rate and ensure the smooth blasting effect, a calculation method for the energy consumption of cavity expansion is established based on the theory of rock breaking and fracture mechanics. The blasting test is carried out with the cast-in-situ plain concrete model, and the law of variation of blasting size, cavity volume and energy consumption with the decoupling coefficient K is obtained. The research results show that the diameter of cavity expansion is 3.43 cm in the process of increasing from 1.25 to 3.375, the average cavity length is 12.7cm, and the cavity volume ranges from 52.98cm³ to 155.56cm³. The expanding-cavity energy consumption is between 0.58kJ and 1.70kJ, and its proportion in the total explosive energy is between 2.9% and 8.5%, and the volume and energy consumption of cavity expansion decrease with the increase of K. K has the most obvious effect on reducing the energy consumption of cavity expansion when it reaches 3.0~3.5, and then tends to be stable; With the increase of K, the degree of pulverization of the rock on the hole wall by explosive energy is reduced so that the volume of and energy consumption of the cavity are reduced. The results of model test and theoretical analysis are mutually verified. It is indicated that a large decoupling coefficient can be selected to reduce the energy consumption ofcavity expansion in a certain range.

The accurate positioning of cleaning area is one of the key factors to reduce the cleaning area and improve the cleaning efficiency.In order to solve the problem of positioning an irregular paint layer area in laser cleaning applications,a outlines regularization method based on grid filling is proposed.The paint layer region image is obtained by the K-means image segmentation method based on L×a×b× color space.The paint layer region image is rasterized to determine the external raster vertices of the paint layer region,and the 4-neighborhood boundary tracking algorithm is used to search for the external raster vertices of the paint layer region in accordance with certain judgment criteria.The searched raster vertices are sequentially connected as regularized shapes,the redundant points and self-intersecting parts of the regularized shapes are deleted,and the regularized shapes with different raster widths are evaluated by using the intersection over union (IOU) as an index,and the raster width corresponding to the largest IOU at the ratio of the area of a single raster to the area of lacquer layer region ranging from 0.025 to 0.04 is selected,and the right-angled polygons generated by this raster width are considered to be the optimal regularized shapes.The regularizations of various contours are compared and analyzed.The results show that the right-angled polygon obtained by the contour regularization method based on raster filling can more accurately express the contour shape of the paint layer and reduce the cleaning area compared with the minimum outer rectangle.

Stable and high-precision localization is a prerequisite for realizing the cooperative autonomous navigation of unmanned ground vehicle (UGV).LiDAR simultaneous localization and mapping (SLAM) often fails to achieve the precise localization in scenarios lacking geometric features,such as corridors,tunnels,and deserts.Therefore,a leapfrog cooperative LiDAR SLAM degradation correction method is proposed for UGVs.This method is used to estimate the normal vector of each feature point in the current frame,and a LiDAR SLAM degradation detection algorithm is devised.When the degradation of environment is detected,the ranging information about two unmanned vehicles is utilized to correct the degradation in LiDAR SLAM.Finally,the locating results are further optimized in the pose graph.Testing on two self-built UGV platforms reveals that the proposed method achieves better mapping performance compared to the current famous LiDAR SLAM methods,demonstrating its significant capability to enhance the locating performance of LiDAR SLAM in degraded scenarios.

The velocity measurement of bullet at the muzzle is easily affected by the complex environment such as smoke and dust,which can not correctly interpret the characteristics of bullet and calculate its muzzle velocity.A saturation photoelectric detection technology is proposed.The theoretical simulation shows that the minimum transmittance of 637nm laser is about 4.2% under the model of rapid diffusion of smoke in a small space,and the circuit photocurrent amplification is 10000 and 15000 times,ensuring that the circuit remains saturated at the lowest transmittance.A high-precision bullet characteristic time extraction algorithm based on the characteristic waveform of bullet is designed according to the output waveform of the circuit.The saturated photoelectric detection technology can be used to effectively measure the time relationship between the bullet and the smoke passing through the laser beam under the interference of strong muzzle smoke,and the high-precision characteristic time extraction algorithm based on the time of bullet and tail crossing the target can accurately calculate the time of bullet crossing the target.The combination of the detection technology and the extraction algorithm effectively improves the accuracy of the calculated results of bullet muzzle velocity.The results show that this method can accurately calculate the muzzle velocity value of bullet,and the velocity error of multiple calculations is less than 0.5%.

A rapid longitudinal-lateral comprehensive control trajectory planning method is proposed for the trajectory planning of reentry gliding vehicles under flight time constraints,which is based on the analytical prediction-correction of drag acceleration profile and the iteration of banked reversal points.In this method,the gliding trajectory planning issue is divided into longitudinal and lateral trajectory planning.In the longitudinal trajectory planning,a single-parameter drag acceleration profile corresponding to the flight time is designed.Furthermore,the profile parameter is corrected based on the time-of-flight analytical prediction,so as to complete the design of the reference drag acceleration profile to meet the terminal altitude,velocity and time-of-flight constraints simultaneously.In the lateral planning,a lateral planning method based on the iteration of the double bank reversal points is designed by analyzing the trajectory length and terminal position control mechanism.The longitude and latitude of target point are introduced to iteratively solve the bank reversal times,which can realize the effective control of the flight time by adjusting the trajectory length while satisfying the terminal position constraints.On this basis,a trajectory iterative correction strategy based on time error is introduced to complete the generation of a time-controllable high-precision three-degree-of-freedom reentry trajectory.Finally,CAV-H reentry gliding is simulated as an example to verify the effectiveness,rapidity and adaptability of the proposed method.Compared with the existing time-controllable reentry trajectory planning methods,the proposed method can comprehensively exert the longitudinal and lateral time control ability in the same computational efficiency and accuracy.Moreover,it boasts a wider range of time adjustability and minimizes the number of bank reversal instances.Additionally,it enables the rapid prediction of the time-adjustable range and the boundaries of reachable domain.

The energy supply mechanism of aluminum powders of the aluminized explosives on the underwater explosion load in the long-time sequence combustion process is studied.The comparison experiments of underwater explosions of aluminized explosives and lithium fluoride-containing explosives are made based on the underwater explosion tank.The influence of aluminum powder combustion on the shock wave and bubble load is analyzed.The results show that the combustion of aluminium powders in CL-20-based mixed explosives releases a large amount of energy,increases the peak shock wave pressure and shock wave energy,and reduces the decay rate of the shock wave.Compared with lithium fluoride-containing explosives,the aluminium powder in CL-20-based aluminized explosives has a higher flame brightness due to its afterburning reaction during explosion,and the afterburning reaction time of the aluminium powder is more than 1ms.The combustion of aluminium powder can increase the maximum radius of the bubble,bubble pulsation period and bubble energy.The maximum bubble radius of CL-20-based aluminized explosives is 11.89% higher than that of CL-20-based lithium fluoride-containing explosives,and the first bubble pulsation period is increased by 25.74%.CL-20-based aluminized explosive reaches its maximum radius with a delay of 5.20ms compared to CL-20-based lithium fluoride-containing explosives.In addition,the rate of change of the bubble pulsation radius of aluminized explosives is higher than that of lithium fluoride-containing explosives due to the continuous energy supply of aluminum powder at the bubble pulsation stage.Moreover,the contraction rate of bubble is reduced,and the duration of the contraction phase of bubble is increased.It is shown that the bubble pulsation process of aluminized explosives has “asymmetry” in time.

To investigate the effect of Y₃Al₅O₁₂ sintering additive on B₄C ceramics prepared by hot-press sintering,B₄C ceramics are prepared using B₄C powders as the raw material and Y₃Al₅O₁₂ generated from the in-situ reaction between Al₂O₃ and Y₂O₃ as the sintering additive by hot-press sintering process at the sintering temperature of 1950℃.The effects of different contents of Y₃Al₅O₁₂ sintering additive on the phase composition,sintering properties,microstructure,and mechanical properties of B₄C ceramics are studied.The results show that the grain boundaries in the B₄C ceramics exist in two forms of B₄C-B₄C clean grain boundaries and B₄C-amorphous Y₃Al₅O₁₂-B₄C grain boundaries when the content of Y₃Al₅O₁₂ sintering additive is less than 7 wt.%,and the grain boundaries in the B₄C ceramics exist in three forms of B₄C-B₄C clean grain boundaries,B₄C-amorphous Y₃Al₅O₁₂-B₄C grain boundaries and B₄C-crystallised Y₃Al₅O₁₂-B₄C grain boundaries when the content of Y₃Al₅O₁₂ sintering additive is more than or equal to 7wt.%.When the content of Y₃Al₅O₁₂ sintering additive is 7wt.%,the best properties of B₄C ceramics are achieved by hot-press sintering.The relative density,Vickers hardness,bending strength,and fracture toughness of B₄C ceramics are 96.2%,20.9GPa,455MPa,and 3.53MPa·m^1/2,respectively.The Y₃Al₅O₁₂ sintering additive affects the mechanical properties of B₄C ceramics prepared by hot press sintering by altering their porosity,grain size,and grain boundary characteristics.

In order to meet the requirements of unmanned underwater vehicle (UUV) launching the torpedos,a new underwater ejection scheme with flexible piston rod is proposed.In the ejection scheme,the high pressure in large depth environment is used to drive the piston to launch a torpedo.An interior trajectory calculation model of torpedo launched from the ejection device with flexible piston rod is established.And the model is verified by the principle test in the high pressure water tank laboratory.The different launching schemes in the related literatures are comparatively studied based on the model.The results show that the ejection scheme with flexible piston rod has the characteristics of smooth acceleration without steep change,large effective acceleration stroke and high pipe exit speed compared with the traditional multistage cylinder launching scheme.Compared with the traditional high-pressure gas-driven piston scheme,the high-pressure water-driven piston scheme is simpler in structure without pressure storage device.A piston buffer scheme based on load reduction in steps proposed in the paper has better load reduction effect compared with the traditional throttle hole buffer scheme.And it won’t cause excessive pressure in the reverse cavity of piston cylinder.Under the ambient pressure of 1.1-6.1MPa,the predicted velocity of torpedo exiting tube varies from 7.0 to 11.6m/s,and the rear reference point of torpedo dose not collide with the tube wall.The results verify the feasibility of the underwater ejection scheme with flexible piston rod,which provides the design basis for the further development of ejection device.

Lubricating oil plays a pivotal role in engines due to carrying a wealth of information about the engine state,and is crucial for characterizing the faults of engine.The engine of an armored vehicle is studied,and a fault diagnosis method for the engine is proposed,which leverages the kernel linear discriminant analysis (KLDA) and an improved dung beetle optimization (DBO) algorithm to optimize a back propagation (BP) neural network.The dimensionality reduction of the acquired lubricating oil data is performed through KLDA,and the dimensionality reduced data is taken as input for the BP neural network.The DBO algorithm is then enhanced by integrating optimal Latin hypercube sampling method,weighting factors and Levy flight strategy in order to further optimize the key parameters of neural network.A fault diagnosis model is established to predict the faults in test data effectively.Experimental results affirm the proposed method’s efficacy in rapidly and accurately predicting the faults,providing a scientific basis for the maintenance and repair of engines in armored vehicles.

The existing visual object tracking methods often face the challenge of tracking failure when dealing with complex backgrounds or drastic appearance changes due to relying solely on the initial frame’s single appearance feature.To address this issue,a single object tracking algorithm based on multilayer feature embedding is proposed.To enhance the discriminability of the target’s appearance,a sparse embedded attention encoder is employed to embed the appearance features with high instance distinctiveness.Additionally,an category feature aggregation encoder is utilized to embed the target’s category information,maintaining the compactness within the class when the appearance changes occur.Simultaneously,the predicted historical frame tracking box coordinates are transformed into the embedded target motion trajectory features,providing the tracking algorithm with high-confidence temporal context features.Experimental results demonstrate that the proposed algorithm achieves a success rate of 71.4% and an accuracy of 92.6% in the OTB100 benchmark test.Moreover,it exhibits robust performance on three large-scale public datasets,namely GOT-10K,LaSOT,and TrackingNet,with success rates reaching 64.9%,72.0%,and 78.7%,respectively.The proposed algorithm effectively overcomes the limitations of the existing tracking algorithms,and has the enhanced tracking accuracy and robustness.

Due to the limitations of processing technology,underwater pressure,safe transportation and other factors,the shell thickness and three-dimensional principal size of the acoustic scattering scale model are often difficult to meet the acoustic similarity design requirements,which seriously restricts the conversion of the acoustic scattering characteristics between the model and the actual ship.Therefore,this paper tries to take the acoustic similarity analysis and correction of the local acoustic parameters of the shell as the basis for the correction of the overall acoustic scattering of a target,and proposes a target strength correction method for the non-coordinated scale model of the shell thickness.The single and double-layer spherical shell models and the complex double-shell model are used for the correction and conversion analysis of target strength,respectively,and the variation law of target strength of the scale model caused by the change in the acoustic parameters of complex shell is summarized,and the effectiveness of the correction method is verified.The proposed correction method can be used to estimate a prototype more in line with the actual engineering needs from the target strength of the shell thickness non-coordinated scale model.

The concave honeycomb structure has broad application prospects in the automotive industry,aerospace,biomedical and other fields due to their unique deformation mode,excellent impact resistance and energy absorption properties,and lightweight characteristics.Based on the traditional Concave hexagonal honeycomb structure,a deformation-controllable concave honeycomb structure based on rounded corners enhancement is proposed by introducing a rounded corner design and changing the arrangement of the rounded corners,and a deformation-controllable honeycomb structure with deformation modes of Z and Y shapes is designed and prepared using metal 3D printing technology.In order to explore its impact resistance,its deformation mode and energy absorption properties are analyzed through the quasi-static compression and drop weight impact experiments and the finite element numerical simulation.The research results show that the proposed honeycomb structure achieves controllable deformation mode and has higher crushing stability,and the energy absorption performance of the structure is significantly improved through customized Z and Y deformation modes.For the same structure,the energy absorption performance gradually improves as the fillet radius increases.As the speed increases,the structural deformation mode gradually evolves into an I type collapse,the platform force generally shows an increasing trend,and the energy absorption efficiency gradually decreases.Due to the asymmetric arrangement of the rounded corners,the Z shape structure has better impact resistance than the Y shape structure in most cases.The research results can provide reference for the crashworthiness design of new structures under dynamic impact.

Titanium alloy is commonly used in the skin-skeleton structure of various aircraft rudder and wing components,which has received widespread attention in the weapons,aviation and aerospace industries.In this work,the laser welding deformation of a titanium alloy skin-skeleton structure is taken as the research object.The influence of welding sequences on welding deformation and stress is simulated and investigated using the thermal cycling method,and the welding sequences are optimized.The welding deformation and stress are significantly reduced by adding the turnover process in the welding.The results show that the peak welding stress is reduced by 27.4% from the original 1027.18MPa to 745.30MPa by prioritizing welding in the center area of the skin and adding multiple overturns during welding.The deformations at the feature points P₁,P₂ and P₃ are decreased from the original 0.168mm,0.178mm and 0.198mm to 0.066mm,0.028mm and 0.021mm,respectively,by 60.7%,84.3% and 89.4%.The laser welding deformation of skin-skeleton structure is measured by a 3D laser scanner.Compared with the experimental results,the average error of the calculated results is 9.98%,which verifies the accuracy of the finite element model and the optimized welding sequence.