Loitering munitions integrate multiple functions such as reconnaissance, attack, and assessment. They have low usage costs, flexible combat deployment, and are difficult to detect and intercept. They have significantly enhanced the strike accuracy and ability against targets and have become important weapons in the Russia-Ukraine conflict, attracting great attention from major military powers. This paper summarizes the characteristics of loitering munitions, introduces the development status of major loitering munitions at home and abroad, including the “Switchblade” series and “BattleHawk” series of the United States, the “KUB” series and “Lancet” series of Russia, and the “Hero” series and “Harpy” series of Israel, which have practical applications. It comparatively analyzes the performance parameters of relevant loitering munitions, combs the key technologies in the development of loitering munitions, including power technology, anti-interference technology, modular technology, and intelligent cooperative combat technology, summarizes the development trends of loitering munitions, and on this basis, analyzes the combat applications of loitering munitions and the defense means against them.

In traditional chined waverider design, the adjustment of design profile parameters is complex and the design intuitiveness is insufficient. To address these issues, this paper proposes a chined upper surface design method based on B-spline curves. The design flexibility and convenience are improved by directly adjusting the design profile through control points. The base profile is constructed by means of Bezier curves, the leading edge and lower surface base profile are determined using the osculating cone theory, and the upper surface profile is designed using cubic and quadratic B-spline curves. The reliability of the proposed method is verified by computational fluid dynamics (CFD) methods, and the aerodynamic performance of the chined waverider is analyzed. The results demonstrate that, as the angle of attack gradually increases, the influence of the chined upper surface on aerodynamic characteristics weakens gradually, and the maximum lift-to-drag ratio appears at the angle of attack ranging from 4° to 6°. The proposed method provides a more intuitive profile optimization means for waverider design, which has reference value for engineering design.

In modern warfare, the damage efficacy of high-speed penetration weapons against high-strength targets such as underground bunkers and reinforced structures has become a focal point of research. This paper systematically reviews the research progress on the dynamic behavior of projectile materials, constitutive models, and structural responses under high-speed penetration. It analyzes the mechanisms of strain hardening, thermal softening, and adiabatic shear deformation under the coupled effects of high temperature and high strain rate, and compares the applicability of typical constitutive models such as the Johnson-Cook model. Key factors influencing mass erosion, critical instability velocity, and structural failure during high-speed penetration are emphasized, along with their underlying mechanisms. Additionally, technical approaches to enhance penetration capabilities through material optimization and structural design are explored, providing valuable references for researchers in related fields.

In order to improve the detection ability of infrared small targets in complex backgrounds, a small infrared target detection algorithm based on joint gradient discrimination and adaptive matching was proposed. The suspected target area is screened out for the first time in the image through multi-directional gradient features, and the adaptive model is generated by using the grayscale information in the region for re-judgment. Quantitative evaluation was established for gradient judgment and adaptive model matching, and confidence functions were introduced to evaluate different suspected target areas and screen out suspected targets. In order to enable the algorithm to be applied in dynamic platforms such as UAVs, an embedded system was built to realize the detection of small targets in real scenes by the detection system through infrared camera framing. By testing different public datasets and comparing with WSLCM (weighted strengthened local contrast measure) and TLLCM (tri-layer local contrast measure) algorithms in different complex scenarios, the proposed algorithm has good adaptability, and the recognition rate under different samples is more than 92%. The algorithm is processed by hardware acceleration through the customization of the IP core of the embedded platform and the co-design of software and hardware, and the real-time video frame rate is greater than 30 frame/s, which verifies its effectiveness.

The mechanical properties of high-entropy alloys (HEAs) are studied. For this purpose, a simple and precise testing system is established for determining the equation-of-state parameters of HEAs. Based on the principle of wave impedance matching, a test setup is designed using the pressure comparison method to obtain the shock adiabat data of HEA materials. The acquired experimental data is then optimized through adaptive clustering detection. After optimization, the confidence intervals for the slope and intercept of the experimentally determined shock adiabat are narrowed from [1.66293, 2.03332] and [4.01158, 4.38089] to [1.6461, 1.92734] and [4.15248, 4.27542], respectively, and the coefficient of determination (R²) is improved from 0.9687 to 0.9849. The Hugoniot equation-of-state parameters for the HEA are determined as C₂=4.214km/s and λ₂=1.787.The Hugoniot equation of state established in this study is applicable within a pressure range of approximately 5.86-32.77 GPa, and its extrapolation to higher pressure regimes necessitates further experimental validation to guarantee reliability. The results demonstrate that the proposed system achieves high-precision measurement of equation-of-state parameters for high-entropy alloys hrough the combination of data optimization algorithms with experimental design. This work provides critical technical support for the performance assessment and practical application of such materials under dynamic mechanical conditions.

To improve the resistance of B₄C/Al composite targets against explosively-formed projectile (EFP) penetration, the surface morphology of ceramic strike face is modified. Seven types of B₄C/Al composite targets featuring different protrusion-array structures on the strike face are examined. The processes of EFPs penetrating into the composite targets at 1.5, 1.7, and 1.9 km/s are simulated using LS-DYNA. The evolutions of the projectile's mass, velocity and kinetic energy during penetration are analyzed. The results indicate that B₄C/Al composite targets with protruded strike-face structures exhibit superior penetration resistance and deceleration capability compared with B₄C/Al flat-faced targets. Among them, the composite target with a pyramidal protrusion array provides the best deceleration performance for the simulated EFP.The R3 semi-cylindrical target demonstrates the best protective performance under normal impact. The penetration direction significantly affects the penetration resistance of anisotropic protrusion-array targets and the sinusoidal-structured target has the optimal protection performance under oblique impact.

The penetration tests on three sets of geometrically similar projectiles with scaling ratios of 1/1, 1/2 and 1/3 are carried out to investigate the size effect of penetration depth of projectile into concrete media. A calculation method for penetration depth with the projectile diameter coefficient as a variable is proposed, and a conversion coefficient model that takes into account the scaling ratio is established for penetration depth. The dynamic strain rate of material in the projectile-target contact zone during the penetration processare quantitatively analyzed through numerical simulation, and the values of penetration depth conversion coefficients under different scaling ratios are ultimately determined. The results show that the size effect exists in the dimensionless penetration depth between the prototype projectile and the model projectile, which arises from the difference in the average strain rates of target in the tests with different scaling ratios. The strain rate increases with the increase of penetration velocity and the decrease of projectile diameter, and does not conform to the geometric similarity scaling relationship. The established penetration depth conversion coefficient is correlated with the target strain rate and the scaling ratio. This conversion coefficient not only quantifies the influence of the material strain rate on the size effect, but also clarifies the mechanism of the size effect of penetration depth in concrete media from a mechanistic perspective.

To address the issue of small penetration aperture in traditional shaped charge warheads against underwater single-layer targets,a novel W-shaped shaped charge structure design method based on the inner cone angle α and outer cone angle β is proposed.The influence of liner structure on jet formation and damage effectiveness is studied.The ratio μ of the lengths of the inner and outer sides of the liner is defined a-s a characterization parameter affecting the underwater formation and damage performance of the W-shaped liner.The influence of the ratio μ on the formation and initiation points on the formation and penetration of annular jet are analyzed through numerical simulations.The results indicate that,the annular jet converges excessively toward the axis when μ is 0.22,forming an explosively formed projectile (EFP).When 0<μ<1,the outward expansion trend of the annular jet gradually intensifies,the head-to-tail velocity difference decreases,and the jet stability improves with the increase in μ.When μ>1,the slug at the jet tail is reduced,the jet head expands,the head-to-tail velocity difference increases,and the jet stability significantly decreases under the influence of the external water medium with the increase in μ.Additionally,the increase in the number of initiation points induces necking at the head of the annular jet,which has little impact on penetration and aperture formation under the condition of small standoff distance.The large-area damage of annular jet to the target plate can be achieved by optimizing the combination of the inner and outer cone angles of the liner to adjust the value of μ and the number of initiation points.

The influence of grooved structure on the penetration and damage performance of copper-aluminum/polytetrafluoroethylene (Cu-Al/PTFE) energetic composite liner is investigated.A energetic composite liner with large cone angle (140°) is numerically simulated and experimented,and the damage effect of a pre-grooved energetic composite liner on concrete target is examined.A comparison shows good agreement between the experimental and simulated results.The influences of groove structure parameters such as groove width,depth,and wall thickness ratio on the damage effect of jet are further analyzed.The results show that the groovee structure has an effect on the distribution of jet energy between radial expansion and axial penetration.The smaller widths,shallow depths and narrow spacing of grooves are conducive to the radial hole enlargement,whereas the larger widths,greater depths and wider spacing of grooves enhance energy concentration,thereby increasing the penetration depth.Additionally,the number of grooves and the wall thickness ratio significantly influence the stability of jet and the ability of reactive materials to follow up.In particular,a wall thickness ratio of 1∶1 between the copper and reactive material layers yields favourable jet formation and damage performance.

To ensure that the UAV (unmanned aerial vehicle) can accurately locate the reconnaissance target, image matching plays a critical role in the localization process, and its performance directly affects the localization accuracy of the targets. However, UAVs often encounter challenges when performing reconnaissance tasks in complex environments, especially when background information is similar, which hinders the effective elimination of a large number of outliers. This paper proposes an improved RANSAC (random sample consensus) algorithm. First, an initial dataset is constructed using triplet relationships. Based on these relationships, an improved strategy for selecting initial data is proposed, which reduces computational costs and improves matching accuracy through neighborhood-based geometric consistency and triangulation. Second, a data subset refinement strategy is introduced to further enhance the algorithm's sampling performance. Finally, the proposed algorithm is compared with other advanced algorithms on UAV reconnaissance image datasets featuring complex environments and similar background information. Experimental results demonstrate that the proposed algorithm achieves higher computational efficiency and correct matching rates, thus improving both the computational speed and localization accuracy, providing a novel method for UAV reconnaissance target localization.

Aiming at the path planning problem of UAVs in complex scenes,this paper proposes a coupling algorithm that uses the improved A-star algorithm and the improved dynamic window method for path planning,so that the UAV has the ability to avoid static and dynamic obstacles.In terms of global planning,by improving the evaluation function of the A^* algorithm,a path planning algorithm that does not rely on the obstacle expansion map is proposed,so that the UAV can plan a safe path against a priori static obstacles.In terms of local planning,an evaluation function for handling dynamic obstacles is added,so that the UAV has good obstacle avoidance capabilities when facing high-speed dynamic obstacles.Aiming at the problem of too many inflection points on the path resulting in frequent acceleration and deceleration after improving the global planning algorithm,a redundant inflection point deletion strategy was proposed.The simulation results show that compared with the traditional algorithm,the improved algorithm has better obstacle avoidance ability and shorter driving trajectory,which verifies the practicability of the algorithm.

The direction-finding performance of guidance and direction-finding system is limited with a small-aperture array and the array configuration optimization is a key step to improve system performance due to platform resource constraints such as weight,volume,and deployment space.In this paper,the Cramér-Rao bounds (CRBs) for wideband two-dimensional direction-of-arrival (DOA) estimation of both scalar arrays and polarization-sensitive arrays are derived,and a CRB-based performance evaluation and configuration optimization method for small-aperture arrays is proposed.Firstly,the development of wideband DOA estimation and the typical two-dimensional array structures are reviewed,and the wideband signal models based on subband decomposition are established for both scalar arrays and polarization-sensitive arrays.For scalar arrays,a closed-form expression for the CRB of wideband two-dimensional DOA estimation is provided.Subsequently,a general framework and closed-form expression for the CRB of wideband two-dimensional DOA estimation of polarization-sensitive arrays are derived,and a performance evaluation criterion for two-dimensional direction finding with such arrays is established.Finally,A two-dimensional array configuration optimization method based on the derived CRB is proposed by considering the constraints ofguidance system platform on the number of array elements and array aperture.The quantitative optimization of array layout can be achieved by constructing an array configuration optimization set and conducting theoretical performance evaluation.The research results provide theoretical support and technical guidance for the array configuration design and performance evaluation of small-aperture guidance direction-finding systems.

As unmanned combat systems have increasingly emerged as a critical asymmetric warfare approach in naval operations, small unmanned surface vehicles (USVs) have become focal points of international military competition.While leading military powers are actively developing and equipping USVs while advancing countermeasure technologies, significant challenges remain. This study adopts the offensive-defensive dynamics of USVs as its analytical framework, examines the evolving nature of naval warfare in contemporary conflicts, and systematically analyzes the technical specifications, operational tactics, and defensive countermeasures associated with advanced small USVs deployed internationally.Building upon the concept of the USV closed-loop kill chain, we proposes an integrated anti-USV defense architecture that synergistically combines radar surveillance, electronic warfare (reconnaissance/jamming), infrared/electro-optical detection, multi-domain interception systems (air-to-surface/underwater), and asymmetric countermeasures. The findings establish a theoretical foundation for advancing USV-related technologies while providing actionable insights for the development and tactical deployment of next-generation counter-USV systems.

In view of the problem that the infrared imaging for terminal guidance of rockets is affected by aerodynamic heating, the temperature response of the sapphire infraredwindow structure caused by aerodynamic heating is obtained through numerical simulation. On this basis, a wind tunnel test is designed to study the effect of aerodynamic heating on the infrared imaging of the seeker. In the numerical simulation, the aerodynamic heating of the infrared dome along the ballistic trajectory is obtained by combining computational fluid dynamics and engineering algorithms, and the temperature response of the dome structure is obtained by the finite element method. In the wind tunnel test, measuring points are installed on the inner surface of the dome to obtain the temperature response of the dome structure under aerodynamic heating, and the imaging effect of the infrared seeker under aerodynamic heating is tested. The research results show that, under aerodynamic heating, the highestinner wall temperature of the dome is 148 ℃, which is much higher than the target temperature of 27 ℃. Based on the sapphire window temperature, it needs to reduce the integration time of the infrared detector and adjust imaging algorithmto perform image correction.

In response to the application requirements of inertial navigation systems for quartz flexible accelerometers with a large range and high dynamic flight accuracy, a new structure for the pendulum assembly of quartz flexible accelerometer is proposed to enhance the range and second-order coefficient of accelerometer. A mathematical model of the designed pendulum assembly is established to derive the scale factor of the designed accelerometer. The design range of accelerometer is theoretically calculated based on the load capacity of servo circuit. The optimization method for the second-order coefficient of the accelerometer is analyzed, and an optimization scheme for the second-order coefficient is provided. Through numerical simulations, it is verified that the deflection and stress of the designed pendulum assembly under full-scale conditions meet the design requirements. Experimental results show that the quartz flexible accelerometer achieves a range of 110g and a second-order coefficient better than 5 μg/g², thereby improving the dynamic application capability of quartz flexible accelerometers in inertial navigation systems.

It is an important direction for the future development of UAV military field to coordinated attack of multi-UAV to accomplish specific strike tasks. Aiming at the problem of coordinated attack of multi-UAV, a typical confrontation scenario is constructed. The unmanned aerial vehicle cooperative attack problem is modeled as a decentralized partially observable Markov decision process (Dec-POMDP), and a unique reward function is designed. The multi-agent deep deterministic policy gradient (MADDPG) algorithm is used to train the attack strategy. Monte Carlo method is used to analyze the simulation experiment, and the results show that after the training of the multi-agent reinforcement learning algorithm, the completion rate of the UAV cooperative attack task reaches 82.9% in specific confrontation scenarios.

In recent years, the development of large-scale low-earth-orbit (LEO) satellite constellations has been progressing at an unprecedented pace, and their potential applications in the military domain have become increasingly prominent. These constellations possess unique technical characteristics that enable them to restructure the traditional kill chain, effectively addressing core challenges such as reconnaissance delays and insufficient cross-domain coordination in precision strikes. This paper provides a comprehensive introduction to the development of large-scale LEO satellite constellations and delves into their application directions in the military field. By examining typical combat scenarios from the perspectives of the kill chain and kill web, this paper analyzes how large-scale LEO satellite constellations can enhance the construction of the kill web and facilitate the closure of the kill chain in the field of precision strike. The findings of this study offer valuable insights and references for the construction of a global and systematic kill web based on a large-scale low-Earth-orbit constellation. Moreover, this research holds significant reference value for the development of China’s anti-access/area denial (A2/AD) strategic system architecture. As related projects in our country continue to advance steadily, the exploration of the potential of large-scale LEO satellite constellations in enhancing military capabilities becomes even more crucial.

The visible light image of the rocket emitted at night provides clear texture and color information of the tail flame, while the infrared image offers a complete outline of the arrow body and saturated tail flame. Due to the high complementarity between the two, fusion of visible and infrared images has been a focal point in research aimed at enhancing rocket imaging observation capability at launch sites. This paper introduces a fusion method for nighttime visible and infrared images based on rocket tail recognition and region reconstruction. In this method, the invariant features of the arrow body are used to search for the tail secant line, which is then utilized to divide the region. Finally, a fusion strategy for region reconstruction is employed to merge the infrared arrow body with the visible tail flame. This approach effectively prevents loss of feature information from both types of images and maximizes semantic information fusion. Experimental results demonstrate that compared with current main image fusion methods, our proposed method significantly improves fusion effectiveness while better preserving feature details from both visible and infrared source images. Furthermore, it achieves quasi-real-time effects within a short time frame.

Infrared tracking of small and weak targets is an indispensable technology in anti-unmanned aerial vehicle systems. The small size, low contrast, strong maneuverability, and complex background of unmanned aerial vehicles pose lots of difficulties and challenges to tracking algorithms. To address these issues, an improved visual tracking Siamese fully convolutional classification and regression network (ISIAMCAR) is proposed. Initially, a side window filter is incorporated into the model to enhance the feature extraction capabilities of the deep convolutional neural network for small targets within the original network. Additionally, a location-aware module is integrated into the network to maintain the efficient propagation of deep convolutional features for infrared weak and small targets detection and tracking, thereby improving the classification and regression capabilities for infrared small and weak targets. The visual and numerical experimental results demonstrate that our approach outperforms existing state-of-the-art methods, achieving the highest tracking success rate and precision rate. Furthermore, the typical computational time of our proposed model has been thoroughly tested, and the ISIAMCAR network we designed meets the real-time requirements of the target tracking task in the anti-unmanned aerial vehicle system. Finally, ablation experiments have been designed and conducted to verify the effectiveness of the side window filtering and location-aware module in the proposed model.

This paper investigates the guidance of missiles striking the maneuvering targets in three-dimensional (3D) space in the presence of model uncertainties, unknown target maneuvers and terminal impact angle constraints. An impact angle control guidance law based on nonsingular fixed-time sliding mode control (NFTSMC) and fuzzy logic is proposed. First, a 3D missile-target relative motion model is established, and the terminal impact angle constraint control is reformulated as a line-of-sight (LOS) angle tracking problem. To overcome the singularity commonly encountered in conventional terminal sliding mode control, a nonsingular fixed-time sliding mode surface (NFTSMS) is constructed, and an auxiliary function is introduced to guarantee singularity-free controller design. In addition, a fuzzy logic system (FLS) is incorporated to online approximate the lumped disturbances induced by target maneuvers and model uncertainties. The fixed-time stability of all signals from the closed-loop system is strictly proved based on Lyapunov stability theory. The simulated results show that the proposed guidance law can achieve the accurate interception of maneuvering targets under different initial conditions. Compared with the existing fixed-time sliding mode guidance strategy, it has significant advantages in suppressing the control command chattering and improving the line-of-sight angle tracking accuracy.

To address the low decision-making efficiency and poor practicality of weapon-target assignment (WTA) in modern air-defense operations,a multi-objective WTA model that comprehensively considers four performance metrics including ammunition consumption,operational cost,total engagement time,and interception benefit is constructed.Practical constraints such as weapon-ammunition compatibility,ammunition inventory,and damage thresholds,etc,are also taken into account to enhance the battlefield applicability of the model.Secondly,a hybrid heuristic algorithm—hybrid ahaotic quantum particle swarm optimization-variable neighborhood search (HCQPSO-VNS) is proposed to solve the proposed WTA model.In the proposed algorithm,a logistic chaotic mapping is employed to improve the quality of the initial population,the quantum particle swarm optimization (QPSO) is utilized for global search; and the variable neighborhood search (VNS) with multiple neighborhood structures is integrated for local optimization to avoid premature convergence.Simulated results demonstrate that the proposed algorithm converges to a high-quality feasible solution within very few iterations.The obtained assignment schemes satisfy the expected lower bounds for damage,weapon-ammunition compatibility,and other constraints,while achieving an effective balance among four performance metrics.Comparative analysis shows that the overall performance of the proposed algorithm outperforms several mainstream algorithms,and can effectively improve the efficiency and scientific rigor of air-defense firepower allocation decisions.Meanwhile,as the problem complexity increases,the proposed algorithm retains high optimization efficiency and acceptable computational load,demonstrating favorable scalability.

Aiming to address the issues of low lift-to-drag ratio during the gliding phase, poor maneuverability in the passive flight segment, and limited range inherent in traditional canard-layout field rockets, this study analyzes the aerodynamic characteristics of canard-layout rockets. Based on these findings, a novel aerodynamic layout design scheme is proposed for high-performance long-range rockets with a large aspect ratio. Simulation results indicate that at an 8° angle of attack, the aerodynamic efficiency of the canard control surfaces decreases sharply with increasing Mach number, while their contribution to the projectile's lift force increases significantly. When Ma ranges from 4 to 6, the proportion of lift generated by the control surfaces is less than the drag they produce, resulting in negative aerodynamic gains for the entire projectile. In contrast, the body lift accounts for 61%~67% of the total lift. Based on these insights, a tail-controlled high-performance rocket layout with a large aspect ratio is proposed. This design reduces the number of inefficient control/wing surfaces during high speed flight and enhances balance and maneuverability through optimized rear-positioned control surfaces. Numerical simulation results confirm that the tail-control layout substantially improves the rocket's lift-to-drag ratio, provides a larger usable angle of attack, and enhances overall maneuverability, making it a more suitable configuration for long-range rockets in hypersonic flight.

To optimize the aerodynamic shape of small-caliber super-high-speed armor-piercing projectiles, a research on the influence of the shape parameters of the armor-piercing projectiles on the aerodynamic characteristics and flight stability was conducted. Based on three key factors, namely the sweepback angle of the tail fins, the aspect ratio, and the overall length-to-diameter ratio of the entire projectile, multiple improvement schemes for the aerodynamic shape were put forward. A numerical calculation model of the hypersonic flow field was established by using Fluent to explore the variation patterns of the aerodynamic parameters and flight stability. Through comparative analysis, the superior aerodynamic shape scheme was acquired. The results suggest that the zero-lift drag coefficient is reduced by 5.3%~26.9% after the improvement, and the lift coefficient and static moment coefficient are better than those of the original projectile. The static stability reserve amounts to 11%~29%, and the logarithmic attenuation rate of the amplitude ε equals 34.39%, meeting the stability requirements of tail-fin-stabilized armor-piercing projectiles. Meanwhile, the firing altitude, velocity drop, and oblique range at each firing angle have witnessed significant improvements compared to the original projectile. This offers a certain reference for the shape optimization design of small-caliber super-high-speed sabot-armor-piercing projectiles.

Random vibration screening test is essential for the development of on-missile electronics used in steering gear control drivers, whether the structure of the vibration test fixture is reasonable plays an important role in the accuracy and reliability of the test. To combat some problems such as excessive mass, poor vibration transmission characteristics and so on existing in this kind of vibration test fixture, a method of optimization design by comprehensively using topological optimization, parameter optimization combined with pre-stress modal analysis and frequency response analysis is proposed. Taking a typical vibration test fixture as an example, the basic shape of fixture structure is obtained by topological optimization, and the interface location between the fixture and the vibration table is adjusted by parameter optimization according to the actual working conditions and design principles. While the mass of the optimized fixture is reduced by 32.65%, the natural frequency and the quality factor of the response point did not deteriorate significantly and the vibration transmission characteristics meet the design requirements. Through the physical random tests, the power spectral density response curve of the test point is basically consistent with the acceleration amplitude-frequency curve of the simulation response point, which proves the accuracy of the finite element model and the effectiveness of the comprehensive optimization design, and provides references for the subsequent design of the vibration test fixture for missile-borne electronics.

To address the challenge of effectively analyzing the dynamic response of structures under high-speed impact, this paper proposes an adaptive sensitivity analysis method based on the Material point method (MPM) and sparse solynomial chaos expansion (sPCE) to address uncertainties in structural response. First, MPM is used to simulate the high-speed impact process of projectiles, revealing the evolution pattern of structural strain energy over time. Second, a surrogate model is developed using the sPCE method to analyze key parameters influencing the dynamic response of the structure and to assess the sensitivity of each parameter to structural behavior. Finally, the proposed method is validated through numerical experiments on projectile impact and penetration scenarios. The results indicate that the intrinsic material properties play a dominant role in structural response, and the coupling effects of impact angle and velocity exhibit strong nonlinear characteristics under specific conditions. The findings provide important theoretical support for optimizing structural design under high-speed impact and demonstrate the accuracy and practical value of sPCE in sensitivity analysis.

Image semantic segmentation assigns labels to each pixel in the target category based on the “semantics” of the image scene, distinguishing different types of things in the image. Existing semantic segmentation methods based on DeepLabV3+ have high computational complexity and large memory consumption, and it is difficult to fully utilize multi-scale information when extracting image feature information, which may lead to the loss of detailed information and reduce segmentation accuracy. The outdoor environment has a variety of targets, large changes in lighting conditions, and obstructions, which increase the difficulty of scene understanding and object recognition. Therefore, an improved DeepLabV3+ network outdoor multi-target scene segmentation method is proposed, with the improved MobileNetv2 as the model backbone. The ECAnet channel attention mechanism is applied to low-level features to reduce computational complexity and improve target boundary clarity. After the ASPP module, a polarized self-attention mechanism is introduced to improve the spatial feature representation of the feature mPA. The improved model has an average intersection to union ratio (mIoU) and average accuracy (mPA) of 69.5% and 82.35% on the outdoor dataset Standford BackgroundDataset, respectively. Compared with the original DeepLabV3+model, it has improved by 5.2% and 4.5%. The addition of new modules has no significant impact on the running time, effectively improving the inference efficiency and accuracy of the model.

In order to study the spinning projectile stability of coning motion,provide the dynamic equation around the center of mass in quasi-body coordinate system and establish the short period dynamic model.Derive the equations of coning motion represented by Euler angles.Obtain the analytical solution by introducing complex angle of attack.Analyze the stability of coning motion with initial disturbance in two cases:One situation is ignoring the Magnus effect and damping effect,only analyze the coning motion influenced by the gyroscope effect and the aerodynamic static moment.Another situation is the coning motion adding in the Magnus effect and damping effect,analyze the stabilization and provide the conditions for coning stability.Finally,summarize the influence of rotating speed and statically stability acting on the dynamic stability of coning motion based on theoretical analysis and simulation results:For the low-speed rotating missiles,only under the influence of gyroscopic effect and aerodynamic static moment,a statically stable aerodynamic shape is necessary to achieve coning stability.Statically unstable missiles can only achieve dynamic stability by significantly increasing the rotating speed.Considering the Magnus effect and damping effect,the convergent coning motion of the statically stable missiles with low rotating speed is the easiest to achieve.The divergence of coning motion may occur result in the increasing of rotating speed.For the statically unstable missiles,it is necessary to select the moment of inertia reasonably and increase the rotating speed to achieve dynamic stability.

In response to the limitations of traditional prescribed performance in controller design, this paper studies a non singular terminal sliding mode fault-tolerant control method that improves prescribed performance. Aiming at the problem of control singularity caused by the convergence speed of the performance boundary of traditional prescribed performance being too fast and the tracking error exceeding the performance boundary, an improved performance function and an improved error conversion function are proposed. The converted error is introduced into the design of a non singular terminal sliding fault-tolerant controller. Simultaneously utilizing an extended state observer to accurately estimate unknown external disturbances in the system, and compensating them to the controller to solve the problem of partial failure of the actuator and suppress elastic effects, in order to obtain preset transient and asymptotic steady-state performance. The stability is proved through Lyapunov theory analysis. Finally, the effectiveness and superiority of the proposed fault-tolerant control method are verified through simulation and comparative experiments of the attitude system of air breathing hypersonic vehicle.

A two-stage cooperative trajectory fast planning method based on pseudo-spectral method and sequence convex optimization is proposed for the cooperative guidance problem of hypersonic gliding vehicle. First, a hypersonic gliding vehicle dynamics and kinematics model is established. Secondly, considering multiple path constraints and terminal constraints in the re-entry gliding section, the pseudo-spectral method is used to solve the complex kinematic model of the vehicle offline, and trajectories that satisfy the relevant cooperative indexes are obtained. Then, on the basis of the offline optimized trajectories, an improved sequential convex optimization algorithm is adopted to adaptively update the radius of the trust region, which improves the convergence speed of the algorithm and realizes the rapid planning of the synergistic trajectories under the premise of guaranteeing the accuracy of the solution. Finally, the proposed method is verified by mathematical simulation, and the results show that the method can realize cooperative flight under the state of uncertainty of the initial conditions and model of the vehicle, and the time cooperative error is less than 0.5 s, and the average CPU time of a single trajectory optimization is 3.67 s, which has a good potential for engineering application.

In order to address the problem of difficulty in determining parameters for the Weibull reliability identification test for complex electromechanical products, a Weibull distributed reliability qualification test method based on historical zero-failure data is proposed.The prior information of failure rate is calculated based on the two-parameter Weibull model and historical data, the posterior information of failure rate is obtained by Bayesian statistical theory, and the parameter estimation of Weibull distribution is obtained by least square method, and the reliability qualification test scheme is calculated. The reliability qualification test scheme is designed and compared with the scheme in GJB 899A—2009 through taking a viewing device as an example in this paper. The reliability assessment result of the Weibull distribution for reliability identification test is 9.2% more accurate than that of the exponential distribution. The established Weibull distribution reliability qualification test scheme based on zero-failure data is effective. The overall Weibull distribution test plan is stricter than the exponential distribution test plan, and this method is highly suitable for the design of reliability qualification test plans for electromechanical products. The established Weibull distribution reliability qualification test plan based on no-failure data is effective and can be directly applied to the formulation of reliability qualification test plans for complex electromechanical products.

The modal parameters of large-scale missile and rocket systems are generally obtained through finite element analysis and ground vibration tests.Due to the inability to simulate the time-varying characteristics of the system under flight conditions,it is usually necessary to conduct finite element modeling or ground vibration tests based on several characteristic conditions,and finally obtain the flight modal frequency through numerical interpolation.The repeated modeling and ground testing process is time-consuming and laborious.This article presents a fast prediction method for time-varying modal parameters of missile and rocket,which is based on the concentrated mass beam model and finite element method.This method introduces time-varying mass matrix,stiffness matrix,and damping matrix to describe the time-varying system,which can quickly provide the time-varying modal frequency parameters of the system.This method has been validated through ground vibration test and flight test,the results show that this method can reliably simulate the time-varying modal parameters of the system,effectively solving the pain points of the tedious simulation of time-varying systems.Moreover,the time-varying dynamic response data obtained based on this method can be used for dynamic inverse problems such as the modal consistency analysis and load identification.

Considering the coexistence of both matched and unmatched unknown disturbances in electro-hydraulic valve-controlled erection systems and the inherent dead-zone nonlinearity in hydraulic valves, a novel nonlinear controller is proposed. First, a nonlinear mathematical model of the electro-hydraulic erection system is established based on the electromechanical-hydraulic coupled dynamics of the system and the dead-zone characteristics of the hydraulic valve. Second, based on the full-state feedback condition, the aforementioned mathematical model is reformulated, and an adaptive extended state observer (AESO) along with a disturbance observer (DO) is designed to estimate both matched and unmatched disturbances, and effectively suppress the observation peaking phenomenon. Third, a novel active disturbance rejection controller is developed leveraging the aforementioned AESO and DO to achieve feedforward compensation for disturbances. Simultaneously, to address the inherent "differential explosion" issue in the conventional backstepping-based controller design framework, a nonlinear command filter is incorporated. At last, through rigorous analysis based on Lyapunov's theory, it demonstrates the boundedness of the motion errors of the system, observer estimation errors and filter error, and verifies the stability of the controller. A simulation platform is developed to validate the performance of the proposed controller. Comparative results with conventional industrial PID controller demonstrate that the proposed controller significantly enhances motion tracking accuracy of the electro-hydraulic erection system.

In view of the technical difficulties such as unstable and target loss in the process of tracking targets of the image seeker, this paper combines the advantages of reliable and stable control of “people in the loop” in manual tracking mode, and the advantages of automated information processing and simple shooter operation in automatic tracking mode, and innovatively proposes an anti-interference stable tracking fusion design method for image seeker. In the automatic tracking mode of seeker, when occasional abnormalities such as target tracking failure or target loss occur, the tracking point is slightly corrected through manual adjustment, which not only reduces the operation pressure of the shooter, but also maximizes the reliability and anti-interference ability of automatic tracking under complex backgrounds, bad weather and other conditions. At the same time, based on the principles of intervention continuity, correction agility and automatic tracking stationary, the optimal solution for micro-revision interval time and correction cycle is designed using the analytic hierarchy process method. The semi-physical simulation results show that the micro-repair preferred solution can achieve a missile hit rate of 99% under different simulation conditions.

Deep debonding integrated propellant column is a high loading grain suitable for small opening & non-segmented shell single chamber dual thrust engines. In order to conduct a safety assessment of the engine in low-temperature ignition, this article conducted simulation analysis on the integrity of the low-temperature structure of the propellant column based on a finite element model, and studied the influence of the pressure difference between the inner hole and gap of the column on the integrity of the charge structure during low-temperature ignition process. The pressure distribution inside the gap during the ignition process was tested by a simulated engine. The results indicate that the pressure difference between the inside and outside of the propellant has a significant impact on the stress and strain distribution of the grain during low-temperature ignition process. Through experiments, it was found that under low-temperature ignition conditions, the pressure building rate at the tail of the gap is synchronized with the combustion chamber whose max pressure is only 4% lower, while the pressure building in the middle of the gap is delayed obviously, and the maximum pressure difference is about 64% of the combustion chamber. The calculated comprehensive safety factor of the column at low temperature reaches 1.76 indicating high safety and reliability of low-temperature ignition.

A rule constrained dual population co-evolutionary genetic algorithm is proposed to address the problems of large solution space, complex constraints, and easy falling into local optima in joint combat firepower strike task planning. Design a joint firepower strike task allocation scheme with rule constraints based on four types of battlefield combat rules, enhance adaptability to complex constraint conditions, construct a dual population co evolutionary solution based on target sorting and firepower allocation sorting, and improve the performance and convergence speed of the algorithm. Experimental analysis shows that this method can obtain feasible solutions with higher performance, has strong applicability and optimization ability, and can effectively reduce fight cost.

The super caliber canard-controlled projectiles have a large front diameter,a forward center of gravity,a variable diameter cross section in the middle of the projectile body,and more complex surface gas flow.In order to study its aerodynamic characteristics,the lifting resistance characteristics,pitching moment characteristics and the surface flow field distribution of the projectile were analyzed by numerical calculation.The results show that the lift-drag ratio increases first and then decreases with the increase of attack angle,and the lift-drag ratio reaches the maximum when the attack angle is 8°.The pitch moment provided by different parts of the projectile body is studied in different areas.It is found that the forward pitch moment provided by the projectile tail takes up more than 30% of the total pitch moment when the attack angle α=0°,elevator angle δ_z=5°,and the stability ratio of the δ_z=10° decreases by 46.8% compared with δ_z=5°.Vortices appear at the variable diameter of the projectile body,which interact with the vortices of the rudder surface and the vortices around the surface of the projectile body,making the gas flow on the surface of the projectile body more complicated and increasing the difficulty of control.

In order to support the selection of materials for the combustion chamber shell of solid rocket motor, an evaluation model for the effective volume, mass, mass ratio, and PV/W coefficient of the cylinder section is established based on thickness. Based on 30CrSiNiMoVA, the variation of the structural characteristics of the cylindrical shell under typical metal and non-metallic material conditions could be obtained through non quantitative derivation analysis and a fast lateral analogy method for the comprehensive benefits of typical material shells is proposed. The results indicate that when calculating various structural features, both metallic and non-metallic materials can be equivalent to the same structural form calculation formula. With the increase of pressure, the change of effective volume and mass ratio of cylinder section caused by material replacement is larger and linear,but it has little effect on the change of mass and volumetric efficiency coefficient of cylinder section; Without considering the material craftsmanship and comparing with metal materials,the use of titanium alloy and fiber materials has advantages in improving the comprehensive performance of the shell,and the advantages of fiber materials are more obvious. When using conventional aluminum alloys, it does not have advantages.

Fragments,as the primary damage elements of blast-fragmentation warhead,have their damage lethality which is quantitatively evaluated through precise testing of parameters such as fragment velocity,spatial distribution and mass characteristics.This paper systematically reviews the latest advancements in the parameter testing technologies for the fragment fields of blast-fragmentation warheads,focusing on comparative analysis under two typical conditions of static and dynamic detonations.In the context of static detonation testing,the principles and features of the contact-type technologies such as net targets,the sectional optoelectronic technologies of light curtains and radar and the 3D reconstruction technologies like high-speed stereovision are compared in detail,and their technological improvements and development trends are elaborated.In the context of dynamic detonation testing,the research achievements in testing methods and simulation modeling at home and abroad are reviewed,and the unique challenges such as detonation point control and spatiotemporal synchronizationunder dynamic detonation conditions as well as the corresponding solutions are thoroughly analyzed.Furthermore,this paper also explores the applications and enabling potential of intelligent algorithms represented by machine learning (particularly deep learning) in the aspects fragment target recognition,trajectory tracking,data fusion,and dynamic explosion parameter prediction.Finally,it offers prospects for the future development trends in fragment field parameter testing technologies,and proposes the need to prioritize high-precision dynamic detonation fragment parameter testing technologies,enhance the 3D reconstruction capabilities and strengthen the integration of machine learning in testing methodologies,thereby supporting the optimized design and damage effectiveness evaluation of blast-fragmentation warheads.

Addressing the control problem of coordinated formation control of leader-follower aircraft, we put forward a missile formation flight controller design method that take the leader aircraft as a virtual target and the missiles fly in formation according to leader's ballistic coordinate system. Firstly, define a relative motion coordinate system in the inertial coordinate frame to establish a relative motion model, which is then transformed into the leader missile's ballistic coordinate system. Secondly, take the position of the follower missile in the line-of-sight coordinate system and the velocity of the leader missile in the inertial coordinate system as state variables and take the acceleration of the follower missile as the control input and then integrate the error between the follower missile's relative position and its desired position, along with the velocity error of the leader missile to design a formation controller using sliding mode variable structure control theory. Finally, simulate and validate a formation structure comprising one leader missile and two follower missiles. Simulation results demonstrate that the proposed method can rapidly and accurately construct and maintain the desired formation.

To investigate the impact of rocket engine gas jet on front cover of a multi-launcher canister during a rocket launching process, a numerical simulation approach was adopted to analyze the flow field after the rocket exits the canister. A gas jet impingement model was established based on the 3D compressible Navier-Stokes equations, the Realizable k-ε turbulence model, and a second-order upwind total variation diminishing (TVD) discretization scheme. The results reveal that as the canister-out height of the rocket increase, the impingement effects on the canister initially intensify and then diminish. Key findings include: At 0.5 m out height, the front cover bears the lowest surface temperature and lowest pressure, with minimum impact loads. At 8 m height, the front cover bears the highest peak temperature of 1 375 K and highest maximum pressure of 195 kPa. At 10 m height, the impact load peaked at 20.21 kN, followed by a gradual reduction in temperature, pressure, and impact loads as the rocket ascended further. This study provides critical insights for safety design of launcher canisters and thermal protection strategies for launch systems.