Since aircraft tire hydroplaning can be influenced by several factors, and the characteristic of evaluation indexes can be described as fuzzy, random, and discrete, a hydroplaning risk evaluation model based on extension theory and connection cloud was established, so as to quantify the transformation of hydroplaning evaluation indexes among different classification levels. The numerical characteristics of the connection cloud were calculated according to the leveling criteria of evaluation indexes, and the connection cloud within a limited range was generated. The extensible matrix of the connection cloud was built by using certainty degrees. In this way, the final risk level could be obtained on the basis of variable weights, which demonstrated the dynamic connection between elements to be evaluated and risk level. The case analysis data was obtained by the fluid-solid coupling simulation of aircraft tire hydroplaning to make up for the lack of variable conditions in the classic hydroplaning test. The analysis results show that the evaluation conclusions of sample 1 and sample 3 are consistent based on the traditional normal cloud model and extensible connection cloud model. The hydroplaning risk level of sample 2 is given as Ⅲ by using the proposed model in this paper. Therefore, the risk control is considered more restrict under the same parameter condition. The confidence factor of the above sample risk assessment is less than 0.01, and the credibility of the evaluation results is high. The proposed model in this paper provides an alternative method for random-fuzzy and uncertainty analysis involving multiple incompatibility indexes. Hence, the defect of the normal cloud model in simulating the distribution of evaluation indexes within a limited range can be overcome.

An algorithm based on Mobile Vit and inverted gated codec is proposed for retinal vessel segmentation (FMVG-Net), aiming to tackle issues such background noise interference, boundary texture blurring, and challenging extraction of microvascular areas. First, we improve the Mobile Vit module, and realize the double joint feature extraction cleverly in the coding part. Subsequently, we use the multispectral attention module. The module reduces the missing image feature information from the frequency domain dimension, so as to accurately segment the foreground pixel of the vessel. Next, we propose a feature adaptive fusion module to establish the context dependence of vascular texture and improve the sensitivity of vascular segmentation. In order to enhance the precision of retinal vascular picture segmentation, lastly, we have implemented an inverted gated codec module and optimized the codec structure to further collect deep and spatial semantic information. Experiments are performed on the DRIVE, STARE, and CHASE_DB1 datasets,whereby the obtained specificity are 0.9863, 0.9897, and 0.9873, respectively; the accuracies are 0.9709, 0.9754, and 0.9760, respectively; the sensitivity is 0.8109, 0.8010, and 0.8079, respectively.Simulation experiments reveal that this paper demonstrates a superior segmentation effect on eye lesions images, which opens new avenues for the diagnosis of eye diseases.

Under high-speed driving conditions, over-area handover, as a key technology for future 5G-R communication of high-speed railways, is crucial for ensuring driving safety. The next-generation 5G-R wireless communication system of high-speed railways adopts fixed handover parameters, but when the train is running at high speed, it is highly susceptible to the Doppler effect, resulting in low handover success. To address this issue, an improved 5G-R adaptive high-speed railway handover algorithm that took into account the influence of the Doppler shift was proposed. First, the influence of the Doppler shift on the handover success rate was analyzed, and the relationship function between Doppler shift and handover success rate was obtained. Then, the dynamic function of handover over the area considering the influence of Doppler shift was proposed, and three functions, namely cosine, cotangent, and cosecant, were designed to adjust the handover hysteresis threshold and time-to-trigger adaptively. Finally, a quantitative comparison analysis of the handover success rate was carried out for different Doppler shift sizes and different high-speed railway scenarios. The results show that the proposed method can effectively improve the handover success rate over the area, and the handover success rate of the cosine, cosecant, and cosecant functions in the viaduct and mountain areas is better than the comparison algorithm and meets the requirement of quality of service (QoS) higher than 99.5% for the handover success rate of China’s wireless communication system. The research results provide a certain theoretical reference for the evolution of the 5G-R system for next-generation high-speed railways.

A dual path supervision and attention filtering network was proposed to solve the problem of blurry boundary classification of target objects in semantic segmentation tasks of remote sensing images. A supervised boundary extraction module was introduced to increase the channel of boundary information, improve the weight of boundary information in semantic segmentation, and enhance attention to the boundary pixels of the target object. The attention filtering module was introduced to filter out spatial details in shallow networks and abstract semantic information in deep networks through attention maps, discarding redundant information in the network to prevent overfitting. The mean intersection over union of the dual path supervision and attention filtering network on the Potsdam dataset and the Jiage dataset was 85.44% and 86.07% respectively, which increased by 1.24%, 1.28% and 1.54%, 1.27% compared with the suboptimal network MagNet and SAPNet. Experimental results show that the proposed network can more accurately segment the boundaries of target objects.

Morphing aircraft is a new concept vehicle that changes its aerodynamic shape by a morphing mechanism to adapt to various flight environments and mission requirements, which has important application value in aerospace, military reconnoitre, and other fields and is the frontier and hot spot of future vehicle research. In response to the new development needs of wide airspace and wide speed domain of the morphing aircraft, a spanwise bending morphing wing mechanism with a multi-closed-loop space mechanism based on the equilateral Bennett mechanism was proposed. Firstly, the geometric characteristics of the equilateral Bennett mechanism were studied, and the design scheme of the multi-closed-loop morphing wing mechanism based on the equilateral Bennett mechanism was proposed. Secondly, the constrained spiral solution method was used to solve the degrees of freedom of the morphing wing mechanism, and the kinematic model was established based on the D-H coordinate transformation method. Then, the 3D model and virtual prototype of the mechanism were established, and the kinematic simulation was carried out to verify the above model. The prototype was made, and the experiment was carried out. The research results show that the proposed morphing wing mechanism needs only one power source to drive the mechanism movement, with a simple structure and high modularity rate, and the proposed morphing wing mechanism can realize the accurate connection between the parts and the expected spanwise bending action. The research results provide a reference for the basic research and engineering application of morphing wings for new types of morphing aircraft.

Angular contact ball bearings will suffer from fault damage after working for a long time, thus affecting the normal operation of the system. In this paper, angular contact ball bearingswith local defects in the outer ring were taken as the research objects, and the discrimination methods of different local defect contours were proposed. A generalized representation model of time-varying displacement excitation of local defects in angular contact ball bearing was established, and the evolution process of local defects and its displacement excitation mechanism were studied. On this basis, the influence of time-varying displacement caused by bearing defects on dynamic characteristics was explored, and a fault dynamic model of angular contact ball bearings was established based on Hertz contact theory. The correctness of the model was verified by experiments. The results show that the rectangular local defects will eventually evolve into trapezoidal local defects. The changing trend of displacement excitation induced by different defect morphologies is different. Compared with the length of the local defect, the width has a greater influence on the displacement excitation. The research results provide a theoretical basis for bearing optimization design and fault diagnosis.

A dynamic visual simultaneous localization and mapping (SLAM) algorithm based on an object detection network is proposed to address the robustness and camera localization accuracy issues caused by dynamic targets in indoor dynamic scenes. The lightweight network PP-LCNet replaces the YOLOv5 backbone network, and the YOLOv5s with the shortest depth and feature map width are chosen as the object detection network. After training on the VOC2007+VOC2012 dataset, experimental results show that the PP-LCNet-YOLOv5s model reduces the network parameters by 41.89% and improves the running speed by 39.13% compared to the YOLOv5s model. In order to eliminate dynamic feature points from the tracking thread of the visual SLAM system, a parallel thread that combines the enhanced object recognition network and sparse optical flow approach is implemented. Only static feature points are used for feature matching and camera position estimation. Experimental results show that the proposed algorithm improves the camera localization accuracy in dynamic scenes by 92.38% compared to ORB-SLAM3.

Cavity flow oscillation phenomenon would occur during the transiton of ramjet with the inlet entrance unobstructed while its exit blocked. As the flow oscillating dramatically, the flight vehicle would face a crisis of instability attitude controlling and structural failure. Due to the problem of fluctuation flow, the unsteady flow characteristics of a supersonic twin-duct inlet with its exit blocked were studied by wind tunnel test and numerical simulation. The effects of the model scale, Mach number of the incoming flow, and boundary layer suction on the characteristics of oscillating pressure of the inlet were acquired. The results indicate that periodic oscillating flow is observed when the exit of the inlet is blocked. The frequency of the oscillating flow is positively correlated to the acoustic velocity of the incoming flow but inversely correlated to the length of the inlet. The oscillating pressure peak is found to approximate the total pressure value of the incoming flow, and it raises obviously with the increment of the Mach number of the incoming flow. The unsteady flow of the inlet with its exit blocked is approximately simulated by the numerical method adopted in this paper, and the numerical simulation result agrees well with that of the wind tunnel test. Furthermore, unsteady simulation results show that the inlet’s pressure is relieved in the duration of flow oscillation by conducting the boundary layer suction method at the internal contracted region, which results in a shock-on-lip state during the backward movement of the shock wave system. A rise of captured flow coefficient is observed when compared to the inlet without boundary layer suction, leading to a 49.47% increase of amplitude peak while 21.78% descending of the frequency for the oscillating pressure.

To satisfy the requirements on accuracy and efficiency of point cloud registration of damaged compressor blades a algorithm for partition based on features of neighborhood points and the corresponding accurate point cloud registration of aero-engine damaged blades was proposed. First of all, based on the covariance matrix, a multi-step partition model was employed to define the method to divide feature sub-blocks, and thus obtain effective feature regions. Secondly, a stable n-dimensional feature vector was constructed in accordance with the local curvature, the maximum distance between points, and the angle property of the maximum normal vector; then, by introducing the iterative closest point theory, the minimum Euclidean distance between the corresponding points and that from the point to the surface between the corresponding blocks were established. The accurate position correction of the two models was realized. Finally, the unit quaternion algorithm was used to complete the accurate point cloud registration of damaged blades. Experimental results show that the proposed algorithm can achieve point cloud registration on the surface of the point cloud model of damaged compressor blades, significantly improving the efficiency and accuracy of registration. Moreover, the advantages and robustness of the unit quaternion algorithm are verified through the point cloud database of multiple groups of aero-engine damaged blades.

The completion time of the cyclic operation of carrier aircraft usually presents significant uncertainty, which directly affects the development of subsequent mission plans for carrier aircraft by carrier mission planners. To determine the distribution pattern of the completion time of the cyclic operation of carrier aircraft and assist planners in tracking the progress of carrier aircraft mission execution, the time uncertainty analysis method for cyclic operations of carrier aircraft was proposed. Firstly, based on the operation process from dispatching to recycling, the stochastic network model of the cyclic operation of carrier aircraft was established. Secondly, a graphic evaluation and review technique network with Laplace transform as the transfer function was proposed, and the characteristics of its transfer function were studied. It was proven that the stochastic netw ork model of the cyclic operation of carrier aircraft could be solved by utilizing the signal flow graph theory. In addition, the Monte Carlo graphical evaluation and review technique（MC-GERT） method was introduced to simplify the parallel branches with the “AND” relationship in the stochastic network and improve the calculation efficiency of the network solution. By the proposed graphic evaluation and review technique method combined with Monte Carlo, finally, the distribution characteristics of the completion time of the pre-flight preparations, sortie and departure, single wave departure operation, and the periodic operation under the cyclic mode of carrier aircraft were analyzed respectively, which verified the feasibility of the proposed analysis method. The results show that the proposed method can be applied to the online analysis of the planning scheme of the cyclic operation of carrier aircraft and can provide certain scheme implementation expectations for relevant departments of carrier aircraft operation planning.

A ground risk assessment model for unmanned aerial vehicle (UAV) operation was developed in order to guarantee the safe operation of UAVs in low-altitude urban areas. It was proposed that a ground risk map that considered multi-factors and multi-levels be used to define the ground risk associated with UAVs and quantify the degree of risk to the ground from flight operations over different areas. First, the typical characteristics of urban low-altitude and the causal chain of UAV crash-landing accidents were analyzed to establish a quantitative ground risk assessment model under the influence of multiple factors, and analyze in detail the mechanism of the influence of each factor on the ground risk. Second, to determine the chance of the danger happening at various locations, the probability density function of the impact region of a UAV crash-landing was computed while accounting for the UAV’s descending behavior and the uncertainty of the wind parameter. Then, the risk value of each geographic reference unit was calculated by combining multiple layers such as urban area map, population density, shelter effect, obstacles and no-fly zones; then the ground risk map of a large regional scale was generated, which can intuitively show the risk level of different areas of the city. Finally, the proposed method was validated and analyzed based on real examples. The findings demonstrate that there are clear differences in the risk distribution across different urban areas, and that the risk level is strongly correlated with the distribution of population density, the impact of surface shelter, the distribution of ground-level obstacle heights, the designation of the city’s no-fly zones, and the UAV’s flight height. The application of ground-based risk maps in the operation of UAVs will help to identify the level of danger of the areas flown over, avoid high-risk areas, and enhance operational safety in urban low-altitude.

In this paper, a 9-component chemical non-equilibrium model and a perfect gas model were used to study the reentry flow of an inflatable reentry decelerator through numerical simulation. The differences in the calculation results of the two models were investigated. The manifestation of the real gas effect was studied, and the reason for the difference between the real gas effect of inflatable reentry decelerator and that of rigid capsule was explored. The results show that compared to the perfect gas hypothesis, the shock wave position is closer to the wall in the real gas effect. After the shock wave, the air temperature decreases, and the wall heat flux decreases. At 83 km, the specific heat ratio of the gas after the shock wave is higher than 1.4, and the air undergoes a dissociation reaction. At 73 km, the real gas effect is very weak, and the specific heat ratio of the gas after the shock wave remains at 1.4. The air still exists in the form of molecules. The main reason for the difference in real gas strength between inflatable reentry decelerator and rigid capsule in the same altitude range is that the inflatable reentry decelerator has a larger resistance-weight ratio than the rigid capsule. Its speed drops faster after entering the atmosphere, and its speed is lower at the same altitude.

Tooling composites prepreg cured at low temperatures and used in high temperature is one of the most promising materials for forming composite tools to manufacture high-precision composite structures in the aerospace field. Tooling composites prepreg utilizes a specially designed two-step method that consists of low-temperature procuring and high-temperature post-curing in order to realize its low cost and high precision manufacture. The curing stress induced in composite tool manufacturing not only affects the accuracy of the initial tool profile, but may produce stress relaxation in the process of thermal recycling, which further causes the change of tool profile accuracy, and then affects the accuracy of components manufacturing. In this paper, the variation rules and influencing factors of profile accuracy of the complex profile characteristics of carbon fiber reinforced composite tools during solidification and thermal recycling were studied by experiments. The results show that the curing stress in the manufacturing process and the stress relaxation in the process of thermal recycling are the main factors that cause the variation of tool profile accuracy, and the profile deviation tends to be stable with the increase of the number of thermal recycling. Subsequent investigation reveals that the primary causes of the tool profile deviation during the manufacturing process are the interaction between the master mold and the composites tool, resin curing shrinkage, and thermal expansion mismatch-induced curing stress. In contrast, the primary cause of the mold surface deviation during the use stage is the relaxation behavior of the residual curing stress. The research results can help to understand the changes in the structural stability of composite tools. Meanwhile, it is of great significance to reduce the manufacturing cost of composite tools, improving the service life of composite tools and realizing the precise manufacturing of complex and high-precision composite components.

With the development of deep learning techniques, learning implicit field for 3D shape reconstruction has become a heated topic, because implicit field can help networks learn a reasonable and sophisticated reconstruction model than explicit methods. However, there are still some challenges to be solved including lacking semantic information, local detail incompletions and so on. Thus, rather than recreating the entire model from a decoder directly, we first reconstruct the semantic components of a single model using an implicit filed structure based on semantic sections in our paper. Then we aggregate the reconstructed semantic parts together to get the final model. Finally, we test those results on the public 3D shape dataset PartNet and compare them to other cutting-edge single-view reconstruction approaches. It’s obvious that using a semantic part-based implicit field can learn more reasonable shape representations for reconstruction.

To investigate the orientation effect of micro-texture on the sealing performance parameters of the floating ring gas film, a rectangular micro-textured hole was selected as the study object. A theoretical analysis model of textured hydrodynamic lubrication control was developed based on gas lubrication theory, and the finite difference method was used to solve the model. The gas film pressure distribution in the sealing gap was obtained. With rotational speed, pressure, eccentricity, and texture depth as independent variables, this study primarily investigated the influence of texture direction angle on sealing performance parameters. The results indicate that the sealing uplift of the micro-textured floating ring increases with rotational speed, pressure, and eccentricity, while increasing micro-textured hole depth results in a decrease in the uplift. Furthermore, the leakage rate increases as rotational speed, pressure, eccentricity, and texture depth increase. However, an increase in rotational speed has little effect on the leakage rate value. The gas film friction also increases with the increase in rotational speed, pressure, and eccentricity and decreases as the texture depth improves. Rectangular micro-texture plays a crucial role in improving the sealing performance parameters. The rational selection of micro-texture direction angle can effectively enhance the sealing performance of the floating ring gas film. These study outcomes provide a novel idea for enhancing the sealing performance of the textured surface floating ring.

Graphene tag antenna, owing to its low price and environmental friendliness, is a promising alternative to metal tag antenna. However, graphene antennas inevitably suffer from property uncertainties due to the manufacturing process of screen printing and the properties of graphene inks, making their real read range different from the theoretical one. A unique interval-based computation algorithm is proposed to guide the design of graphene antennas with uncertainties considered, as well as to make the read range particular. Firstly, uncertainty parameters are obtained by analyzing the theoretical formulas of antennas. Second, using an interval analysis approach that includes the interval mathematical modeling and the vertex algorithm, the upper and lower bounds of the read range are determined based on preparation experience, with each uncertainty parameter approximated as an interval set. In the end, experiments are carried out to measure the read range of graphene antennas to verify the proposed algorithm. Results show that the proposed algorithm can calculate the read range of graphene antennas effectively. Thus, this research facilitates the practical use of graphene antennas and provides guidance for uncertainty analysis during manufacture.

Unexpected breakdowns of ion thrusters affect their reliability in engineering applications. In order to reduce the frequency of unexpected electric breakdown between shell and screen grid of the thruster, the unexpected breakdown mechanism was clarified through theoretical analysis and experimental methods. By considering the thermal design and the elimination of unexpected electric breakdowns, the shell optimization design and thermal characteristic analysis of three schemes were carried out. A shell scheme verification and heat balance test system was built. The results show that the frequency of unexpected electric breakdowns is significantly higher at high power than at low power. After the single-sided anodized shell on the inner surface is adopted, the frequency of unexpected electric breakdowns of the thruster is reduced from 6.90 times/h before optimization to 0.70 times/h under 5 kW working condition and from 3.3 times/h before optimization to 0.20 times/h under 3 kW working condition. The maximum temperature measured at the electrical connector is 148.5 ℃, meeting the application requirements. The effect of different shell schemes on the temperature change of the thermally sensitive assembly is small, with a temperature difference within 5 ℃, and the maximum error of the thruster temperature test and thermal analysis is less than 10 ℃.

After analyzing the occurrence and the reason behind the power reflux in toyota hybrid system (THS) at high speeds, a novel hybrid electromechanical coupling system scheme that can address the power reflux issue is put forth. The system not only solves the power reflux problem, but also adds an engine direct drive mode and reduces fuel consumption by adding a brake based on THS. By comparing the fuel consumption of the two systems under constant speed conditions, it is found through simulation that the fuel consumption at 90 km/h and 120 km/h can improve by 7.7% and 3.9% after adding brakes. Furthermore, an analysis is conducted on the control strategies employed by the two systems under the two constant speed conditions. The findings indicate that the enhanced system circumvents THS’s power return issue, enhances system efficiency, and optimizes the vehicle's fuel consumption when it employs the engine direct drive mode.

This article presents a comparative experimental study on the flow and heat transfer performance differences between additive manufacturing (AM) rough channels and machined smooth channels for supercritical pressure n-decane. Using a roughness gauge, the surface average roughness of the smooth channel and AM channel were determined to be 3 μm and 11 μm, respectively. The study investigated variations in the friction factor (f) and Nusselt number (Nu) of the two channels under different flow rates, system pressures (3 MPa, 5 MPa, and 7 MPa), and wall heat flux densities (0.29 MW/m², 0.32 MW/m², and 0.35 MW/m²). The results showed that, within the selected flow rate range, both f and Nu of the AM channel were higher than those of the smooth channel, with f being 2.8～3.5 times that of the smooth channel, and Nu being 1.7～2.3 times that of the smooth channel. Increasing the system pressure reduced f of the AM channel, but increased that of the smooth channel, while Nu decreased for both channels. As the wall heat flux density increased, both f and Nu of the AM channel decreased, while those of the smooth channel increased. The rough channel had a comprehensive heat transfer coefficient of 1.2～1.5, showing outstanding overall heat transfer performance and greater effectiveness at low flow rate. The correlations of the Nu and f of supercritical hydrocarbon fuel in additive manufacturing channel are proposed.

The structural integrity of the aero-engine is related to flight safety. Currently, the detection of aero-engine defects based on borescope technology is mainly manually operated. In order to improve detection accuracy and efficiency, an intelligent aero-engine defect detection algorithm by fusing attention and multi-scale features was proposed to assist the borescope detection. In view of the imbalanced distribution of defect classes in original borescope images, a multi-sample fusion data augmentation method based on geometric transformation and Poisson image editing was used to enrich the small sample images, and the defect dataset was constructed. The coordination attention (CA) mechanism was integrated into the baseline network YOLOv5 to emphasize the extraction of defect features and enhance the network’s distinction between defect targets and complex backgrounds. The weighted bidirectional feature pyramid network structure (BiFPN) was constructed in the neck network to achieve a higher level of feature fusion and improve the expression ability for multi-scale targets. The bounding box regression loss function was defined as the efficient intersection over union (EIOU) loss. The fast and accurate location and recognition of defects were realized. The experimental results show that the average precision of the proposed algorithm in detecting defects is 89.7%, 6.3% higher than that of the baseline network. The size of the trained model is only 14.0 M. Therefore, the proposed algorithm can effectively detect the main defects of aero-engines.

In view of the heat flow, overload, and dynamic pressure constraint problems during the reentry process of the lift reentry vehicle, a calculation method of the reentry reachability region under multiple constraints is designed. Using the virtual target method, an optimal control model for the reentry reachability region is developed without the state inequality restrictions. A predictive correction method is designed based on the optimal control without process constraints, which converts state inequality constraints into control inequality constraints. The numerical simulation of the X-33 is completed. The results of the numerical simulation demonstrate that, in contrast to the conventional "soft constraint" method, which depends on the quasi-equilibrium glide conditions, the suggested method is capable of achieving “hard constraint,” meaning that all process constraints can still be strictly satisfied even in the event that the aircraft makes large maneuvers quickly.

In order to improve the quality and reliability of physiological signals, blind source separation and wavelet threshold methods were coupled to propose a multi-spectrum adaptive wavelet signal enhancement method, which was combined with an improved blind source separation method for denoising. To evaluate the effectiveness of the proposed method, the signal-to-noise ratio (SNR) and root mean square error (RMSE) indicators were calculated by using three wavelet transform methods: soft threshold, hard threshold, and adaptive threshold. The results show that the proposed method has strong applicability under the soft threshold, and compared with that under a hard threshold, the enhanced signal under a soft threshold has an SNR improvement of about 44.2%, RMSE reduction of about 28.8%, and a time reduction of about 1.4%. Compared with the adaptive threshold, SNR is improved by about 706%; RMSE is reduced by about 16.7%, and time is reduced by about 3.0%. The original, noisy, and enhanced signals are analyzed and normalized by using a soft threshold to compare the differences in various parameters under a soft threshold. The results show that the enhanced signal has a better SNR, lower RMSE, and shorter processing time. Compared with the original signal under the soft threshold, SNR is improved by about 0.12%; RMSE is reduced by about 2.5%, and time is reduced by about 3.9%, which further verifies the effectiveness of the proposed algorithm and improves the signal quality.

By combining the characteristics of high hovering performance and control efficiency of coaxial-rotor helicopters and the advantage of high-speed forward flight performance of double-thrust-propeller compound configuration helicopters, the coaxial rotor was applied to the double-thrust-propeller compound configuration helicopter. The single rotor was changed to a coaxial rotor based on the configuration of the X3 compound helicopter. To study the influence of coaxial rotors on the aerodynamic characteristics of compound configuration helicopters, a fast trim method was established. On this basis, the aerodynamic characteristics of different configuration helicopters were analyzed. The results indicate that the aerodynamic characteristics of the coaxial-rotor helicopter have good symmetry compared with the single-rotor configuration. In the case of maintaining good high-speed forward flight performance, the coaxial-rotor configuration can significantly improve the hovering and low-speed flight performance. When the hovering efficiency is increased by 6.8%, and the forward flight speed is 100 km/h, the total required power is reduced by 23.1%. The aerodynamic interference of the coaxial rotor to the propeller at low speeds is greatly lower than that of the single rotor.

Unmanned aerial vehicles (UAV) are widely used in low-altitude areas for power inspection, search and rescue, reconnaissance, and other tasks. The detection of obstacles in advance during flight aims to ensure the completion of established tasks. In order to meet the requirements of obstacle detection accuracy and position regression accuracy of UAV flying at low altitudes, a low-altitude UAV obstacle detection method based on improved position constraints and attention mechanism was proposed. The deficiency of position regression loss function was analyzed. On this basis, the loss function of separation scale loss and fusion direction constraint was proposed to optimize the regression process. The improved attention mechanism CBAM proposed a dual attention mechanism to strengthen the feature suppression interference and improve the detection performance. The experimental results show that the proposed method improves by 2.28% on mAP and 2.7% on mAP@0.5:0.95, showing better detection performance of low-altitude obstacles in terms of both detection accuracy and position regression accuracy.

Large flexible aircraft possess large structural deformation with aerodynamic loads, which makes the dynamic characteristics change obviously. Structural deformation predictions are vital to large flexible aircraft design and aeroelastic simulations. Full-order models like the finite element method have low simulation efficiency. Traditional reduced-order model (ROM) obtains high simulation efficiency but requires large amounts of sample data participating in model building. This study investigates machine learning algorithms to build a prediction model to calculate the static deformation of flexible structures considering geometric nonlinearities. The performance of the prediction models is compared under evaluation with root mean square error (RMSE). It is shown that several machine learning techniques can be applied to the prediction of large deformations. Moreover, a new static aeroelastic analysis method is proposed with a large deformation prediction model and non-planar vortex lattice method (VLM) with high accuracy and efficiency In the end, a single-beam flexible wing model is used, and the prediction model and wind tunnel test's static aeroelastic response are contrasted. The results demonstrate that the proposed method has good performance and great practical application value.

Frequency modulated continuous wave (FMCW) Doppler fuze is easy to be interfered with on the battlefield, resulting in an early explosion and loss of damage ability. To improve the anti-jamming ability of FMCW Doppler fuze against information-based jamming and realize the distinction between multiple jamming signals and target echoes, this paper proposed a method of target and jamming signal classification and recognition based on supervised contrastive learning. Firstly, the backbone network was constructed by residual network and self-attention mechanism. Then, the contrastive learning loss function was improved by introducing labels, and supervised contrastive learning was realized. Finally, an intermediate frequency signal was used to build the dataset, and the network was trained by supervised comparative learning, so as to realize the classification and recognition of the target and jamming signal. The simulation results show that this method can realize the recognition of multiple jamming types and target echoes, and the recognition rate can reach 98.7%. In the low signal-to-noise ratio (SNR) environment, the recognition effect is better. In the SNR environment of −18 dB, the recognition rate is still 91.81%, which is higher than the 86.12% recognition rate of ordinary residual networks.

Metal honeycomb cylinder structures frequently encounter a decrease in load-carrying capacity due to axial compression buckling. Extensive research has been conducted on their elastic buckling behavior. However, there are few theoretical studies on their plastic buckling behavior. To address this issue, a theoretical formula was derived for the elastic and plastic buckling critical load of the aluminum alloy honeycomb cylinder under axial compression, drawing upon the equivalent model of the honeycomb cylinder and the Hamilton principle. The simulation analysis and experimentations on the above problem were carried out. Finally, the influence of the honeycomb cylinder dimensional parameters on the critical load was discussed using the theoretical formula. The results show that both the theoretical formula and the simulation model can accurately predict the critical load of the aluminum alloy honeycomb cylinder. In comparison, the theoretical formula is calculated faster. When the cylinder mass changes within the range of ±10%, the influence degree of panel thickness, cylinder radius, core thickness and cylinder height on the critical load decreases successively. The honeycomb cylinder’s critical load exhibits a minor association with the cylinder height, but it exhibits an approximately linear positive correlation with the cylinder radius, core thickness, and panel thickness.

Until recently, object tracking algorithms connected to deep learning have not considered the distinctions between regression and classification branches while employing shallow and deep feature fusion. Both branches used the same fusion feature, which could not meet the different task requirements of each branch well at the same time. According to the relationship between different task requirements of branches and features, a new feature fusion method was proposed for the object tracking algorithm. The channels of different feature layers in the backbone network were adjusted proportionally to form the fusion features suitable for the classification branch and regression branch. To compare and prove the effectiveness of the new feature fusion method, optimization was carried out on the basis of the SiamCAR algorithm. By changing the method of feature extraction and fusion, the accuracy of 2%−3% was improved on the three datasets of UAV123, GOT-10k and LaSOT. The experimental findings demonstrate the efficacy of the novel feature fusion technique, and the framework as a whole achieves good tracking performance at a real-time running speed of 75 frames per second.

By using a cellular membrane capacitor and a potassium ion memristor to characterize the charge storage and memory function of a neuron, this paper modified the classical circuit model and got the mesoscopic circuit model of the Hodgkin-Huxley single ion channel neuron. Based on the classical circuit theory and quantum theory of mesoscopic circuits, the response of membrane voltage of neurons under the sinusoidal excitation was derived. The calculation results show that with the increase in the frequency of the excitation source, the voltage peak-peak value of the membrane of the neuron and its steady-state hysteresis loop area first increase and then decrease. In the classical circuit model, the frequency at which the voltage peak-peak value and the hysteresis loop area reach the maximum depends on the amplitude of the external excitation source, while in the mesoscopic circuit model, it only depends on the circuit parameters of the neuron and is independent of the external excitation, which can better describe the characteristics of neurons. The mesoscopic circuit model of neurons is beneficial to reveal the dynamic mechanism of neural networks and promote the development of brain science theory systems.

The unit load device (ULD) is commonly shaped irregularly to fit the arc shape inside the cabin. This results in efficient loading of irregular ULD at the airport cargo station, which helps to increase the overall efficiency of air freight transportation. Considering a series of actual loading constraints such as cargo volume constraints, a multi-objective loading optimization model for irregular ULD is established with the goal of maximizing the utilization rate of volume, loading cargo quantity, and load utilization rate. Moreover, a heuristic algorithm is constructed based on the principles of three-dimensional space division; the NSGA-Ⅱ algorithm is used to perform operations such as crossover, mutation, and selection. Finally, the algorithm generates a Pareto frontier set, and selects the optimal loading scheme with the greatest volume utilization. The experimental results show that the improved NSGA-Ⅱ algorithm still maintain a high volume utilization rate of 93.84% under strong heterogeneous cargo loading conditions; compared to the initial NSGA-Ⅱ algorithm, the proposed algorithm can increase the volume utilization rate by 13%, indicating that the improved NSGA-Ⅱ algorithm has good adaptability on irregular ULD loading. The research results can provide a reference for irregular ULD loading at airport cargo stations.

In view of the trajectory tracking problem of the forward-tilting morphing aerospace vehicle with time-varying disturbance, a non-singular terminal sliding mode control scheme based on immersion and invariance (I&I) theory and fixed time convergence theory was proposed. Firstly, a time-varying disturbance observer based on I&I was designed by combining dynamic scale factors. Secondly, a segmented fixed-time non-singular terminal sliding surface was constructed, which eliminated the singularity of the sliding mode surface and made the system state converge to any small neighborhood of the equilibrium point within a fixed time, and the upper bound of the convergence time had nothing to do with the initial state of the system. Finally, based on the Lyapunov stability theory, the global fixed-time stability of the system was proven, and the upper bound of its convergence time was given. The effectiveness and superiority of the proposed control scheme were verified in two experimental scenarios. Compared with the traditional control method, the control scheme proposed in this paper made the system converge faster and has better anti-disturbance ability.

The algorithm solves the distributed cooperative task assignment problem with complex constraints among unmanned aerial vehicles (UAVs) with different capabilities in a more flexible way. Better solution accuracy and efficiency can be ensured by each UAV in the distributed task allocation framework by enabling distributed asynchronous computation and asynchronous communication. In order to prioritize significant targets with limited resources and the quickest total journey time, the task assignment principle is based on proximity assignment while taking the target value and schedule limitations into consideration. Simulation results show that this heterogeneous multi-round distributed auction algorithm can obtain good assignment results for any given number of UAVs and targets and UAV combat capability configuration.

The prediction of aircraft taxiing trajectory helps to solve operational problems such as taxiing conflicts and long waiting times at airports, ensuring surface safety while improving service levels and increasing airport throughput. A model is proposed to predict the taxiing trajectory of a surface aircraft based on an attention mechanism that combines gated recurrent units (GRU) and an improved Kalman filter algorithm (IKF). This addresses the current situation where the performance of machine learning models depends on good data sets. In order to better extract data discrepancy features and learn input-to-output mapping relationships, three independent networks of gated recurrent units are first used to capture the future moment motion states and temporal dependencies of the aircraft. An enhanced extended Kalman filter is then fused with the neural network outputs to integrate them into the state prediction and update process, ultimately improving the predicted trajectory sequence accuracy. Finally, the validity of the model was verified using real aircraft taxi trajectories at Lukou Airport. The simulation results show that the proposed model can effectively and accurately predict aircraft taxi trajectories at the field with an overall mean square error of approximately 0.00128. Compared with the single recurrent neural network (RNN), long and short-term memory network (LSTM) and GRU model, the root mean square error (RMSE) is reduced by 72.9%, 54.7% and 39.9% respectively, and the prediction time is 40 ms, which could accurately and quickly predict the taxiing trajectory and provide assistance in reducing the operating load of the airport surface management system.