To systematically investigate the effects of laser power， laser scanning spacing， and laser scanning speed on internal defects and micro-grinding performance of high-entropy alloys， FeCoCrNiAl0.5 high-entropy alloy samples formed by selective laser melting under different combinations of laser process parameters were used. The characteristics of internal defects such as pores and microcracks were analyzed by using metallographic microscopy and scanning electron microscopy. Single-factor tests on micro-grinding were conducted by using micro-grinding force， surface roughness， ground surface morphology， and subsurface damage as evaluation indicators to quantitatively analyze the micro-grinding performance of samples prepared with different laser parameters. The results show that moderately increasing laser power can reduce pores and cracks， improve density and strength， increase micro-grinding force， and reduce surface roughness. Increasing the laser scanning spacing weakens the weld overlap and increases internal defects. Although the grinding force decreases， surface and subsurface damage and roughness are exacerbated. A moderate laser scanning speed is beneficial for suppressing defects， improving density and strength， and thus improving surface quality.
During the preparation of complex-structured ceramic parts using ceramic vat photopolymerization additive manufacturing technology， various types of defects are readily generated due to the inherent attributes of the material system， the layer-by-layer printing principle， and the post-processing techniques. In order to resolve the constraints and damages caused by these defects to part forming accuracy， service performance， and technology promotion， a systematic investigation was conducted on the defect formation mechanisms and control strategies in ceramic vat photopolymerization additive manufacturing. Firstly， two types of defects existing in ceramic vat photopolymerization additive manufacturing were summarized， namely morphology defects （such as dimensional deviation and morphological distortion） and structural defects （such as anisotropy and porosity）， and their complex formation mechanisms were revealed. Then， defect analysis methods based on theoretical modeling and numerical simulation were summarized. In terms of defect control and performance improvement， not only were traditional optimization methods including degassing， slurry improvement， and optimization of debinding and sintering summarized， but also innovative strategies such as “microstructure design” were discussed. Finally， the research directions and development priorities worthy of future attention were summarized and prospected， with a view to providing theoretical reference and practical guidance for promoting the development of high-performance ceramic vat photopolymerization additive manufacturing technology.
In order to study the molten pool and spatter behavior during selective laser melting， a high-speed camera monitoring platform was built. The high-speed camera was used to capture the melting images in the single-channel scanning process under different shielding gas directions and linear energy densities at different exposure times. An image processing method was proposed to process the collected melting process images. The characteristics of spatter under different linear energy density and shielding gas direction were analyzed from the aspects of spatter number， area and angle. It is found that compared with the linear energy density， the direction of the shielding gas direction has a more significant effect on the spatter characteristics. It is recommended to use a lower linear energy density （within an appropriate range） in the selective laser melting process combined with the reverse airflow direction to suppress the spatter.
To address the issue of concave and convex deviations on the surface of laser deposition manufacturing， an algorithm for surface deviation monitoring and feedback control was designed. Surface deviation monitoring was achieved through three-dimensional point cloud reconstruction， coordinate system registration， simplified noise reduction， and deviation value calculation. The feedback control algorithm extracted contour profiles of deviation zones， generating layered compensation trajectories for concave areas and zoned control trajectories for convex areas， thereby implementing feedback control over surface deviation regions. A validation experiment was conducted： the surface deviation monitoring and feedback control programe promptly identified concave and convex deviation zones on the test specimen surface and applied feedback control， significantly reducing surface flatness errors. This demonstrated the effectiveness of the forming surface deviation monitoring and feedback control algorithm， holding significant implications for the advancement of laser additive manufacturing technology.
To tap into the ecohomic benefits and scientific research value of repair techonlogy for aerospace high-strength steel components and meet the development needs of aerospace field，laser additive manufacturing repair technology was employed to repair cracks in 300M high-strength steel substrates. The microstructure， element distribution， static mechanical properties， and fracture morphology of the repaired specimen were analyzed. Results indicate that the microstructure of the repair zone exhibits a gradient characteristic from the top layer to the substrate： Bainite → Martensite + Bainite → Tempered Martensite. The heat-affected zone primarily consists of non-uniform Martensite. After heat treatment， all regions transform into a uniform mixture of tempered Martensite and Bainite. The as-deposited specimen shows higher hardness in the repair zone compared to the substrate， but softening occurs in the heat-affected zone. The as-deposited tensile specimen demonstrates low tensile and yield strength， good plasticity and a quasi-cleavage fracture morphology. After appropriate heat treatment， the strength increases significantly， but plasticity decreases markedly， and the fracture mode transitions to ductile fracture.
Aiming at the problems of local optimum and target unreachability existing in low-altitude UAV path planning under complex obstacle environments，an improved artificial potential field （APF） algorithm was proposed for dynamic path planning of low-altitude UAVs. The proposed algorithm featured three key enhancements： first， the repulsive potential field function was optimized through dynamic adjustment of repulsive intensity based on obstacle distances， effectively resolving target inaccessibility. Second， the gravitational potential field function was refined with dynamic weight allocation to prevent local optima entrapment. Finally， a smooth optimization strategy for resultant vectors was introduced to enhance planning stability and dynamic adaptability. Experimental results demonstrate that compared with the standard APF algorithm， the improved method successfully overcomes both local optima and target inaccessibility issues in complex obstacle environments. This research provides valuable insights for dynamic path planning of low-altitude aerial vehicles.
To investigate the collaborative optimization problem of unmanned aerial vehicle （UAV） distribution center location-allocation in the context of low-altitude urban last-mile delivery， the simulation experiments were conducted on 20 universities within the no-fly zone-exempt areas of Shenyang， which comprehensively considered geographical regional characteristics， UAV performance attributes， and delivery time constraints to address the location-allocation problem within this specific operational context. The characteristics of unmanned aerial vehicle logistics distribution were examined and analyzed the location-allocation problem of distribution centers. Considering specific regional environments and the unique features of unmanned aerial vehicle， a multi-constraint mathematical model was constructed with the objective of minimizing total costs. The elbow clustering method was employed to determine the number of distribution centers， while a simulated annealing optimization algorithm was applied to derive optimal coordinates and allocation scheme under the goal of minimizing total economic costs. The influence of distribution center load rates and capacity constraints on location selection was also considered. Finally， a simulation experiment based on real-world conditions was conducted using Python to obtain the algorithm's optimal solution—namely， the best location-allocation scheme.
To address the practical dilemma of the mismatch between talent supply and industrial demand in the field of "capability-shaping" talent development for the low-altitude economy in the digital technology era， the theories of human-computer interaction， human factors engineering and flow theory were integrated， which constructed a "capability-shaping" talent cultivation framework for the low-altitude economy with aerospace culture dissemination as its core value. Based on the fuzzy set qualitative comparative analysis （fsQCA） method， that discussed the driving mechanism of ability acceptance， learning experience adaptation， virtual platform acceptance， learning commitment， industrial scene docking and policy support perception on the high willingness to accept the training mode. The results show that there is no single necessary condition， and there are three sufficient configuration paths： experience-oriented， platform-driven， and dual-core enhanced， showing multiple concurrent causal characteristics. Based on this， a new paradigm of talent development with five-dimensional core competence as the goal， three-theory integration as the logic， digital virtual platform as the carrier， and aerospace culture as the background was constructed， providing theoretical reference and practical path for the optimization of low-altitude economic talent training.
To reveal the root cause of the sudden vibration-exceedance fault during the 150h endurance test of a certain type of turboshaft engine， the physicochemical inspection was combined with thermal-mechanical coupling simulation to carry out a full-chain mechanism analysis. The results show that severe transient thermal shocks induced by the anti-icing system’s intervention cause typical γ′-phase rafting degradation and accelerated creep elongation in DZ22 alloy blades. This excessive radial growth then triggered intense rotor-stator rubbing， generating an unbalance exceeding 60 g·cm and ultimately causing a sudden rotor vibration exceedance. Larson-Miller model calculations confirm that the extreme thermal load drastically reduces the local creep life from 150 h to approximately 80.5 h， precisely matching the actual removal timing and location. Based on these findings， a sequential control strategy of “reduce operating condition first， then activate anti-icing” was developed. Bench test results demonstrate that this strategy completely suppresses thermal overshoot at the source， eliminates the risk of rotor rubbing， and successfully ensures the safety of subsequent endurance tests for the same engine type.
To address the excessive reliance of existing object segmentation methods on optical flow maps and fully leverage the temporal and spatial correlations inherent in video objects， a two-stage method was proposed that first extracts regions of interest （ROI） using the temporal feature information of objects and backgrounds， and then segments objects using spatial saliency features. Firstly， an optical flow anomaly region detector was designed， which used the optical flow field generated by pixel position changes between sequential images as temporal features to learn the differences in temporal features between objects and backgrounds， thereby detecting object positions and guiding the segmentation scope to the ROI where objects were located. Then， the proposed DC-U2Net was used as a salient object detector to segment salient objects in the ROI， obtain masks， and complete object segmentation. DC-U2Net further improved the accuracy of object masks by introducing a dual-channel RSU module and residual connection paths. The proposed method was tested on two public datasets and achieved competitive results. On the DAVIS2016 dataset， the J-score， F-score， and G-score reached 83.7， 86.7， and 85.2 respectively. Aiming at the characteristics of temporal features and saliency features of video objects， the proposed two-stage video object segmentation method guided by temporal features-guided improves object segmentation accuracy without compromising real-time performance.
