The state of charge （SOC） of lithium-ion batteries is a critical parameter in the battery management system of new energy electric vehicles. To address the issue of insufficient SOC prediction accuracy for lithium-ion batteries under complex operating conditions，an intelligent SOC prediction method for electric vehicle lithium-ion batteries based on the Transformer neural network was proposed. Taking the Nissan Leaf battery as the research object， a charging and discharging test platform for new energy electric vehicle lithium-ion batteries was built to simulate the real energy demands of users and the dynamic changes in real-time energy needs. This platform dynamically adjusted the battery’s charging and discharging strategies， collected multi-dimensional battery data， and preprocessed the data. Then， a SOC prediction framework based on the Transformer model was constructed， which extracted complex time series features through neural networks， achieved high-precision predictions of lithium-ion battery SOC. The experimental results indicate that the proposed method outperforms other networks in prediction accuracy， with a mean absolute error of less than 1.51% and a RMSE of less than 0.48%， validating the effectiveness and accuracy of this method.

For manufacturing enterprise managers，mastering digital leadership adapts to the development of the digital era is the key to promoting the transformation and upgrading of manufacturing enterprises. Based on leadership theory，a path was explored for manufacturing enterprise managers to improve digital leadership，which summarized six antecedents affecting digital leadership，and empirically investigated them using the fuzzy set qualitative comparative analysis （fsQCA） method. The research shows that among all antecedent condition variables，no single element could individually contribute to the realisation of high digital leadership in manufacturing enterprises. There are two paths to achieve high digital leadership capabilities for managers in manufacturing companies. There are four paths to achieve non-high digital leadership capabilities for managers in manufacturing enterprises，which are causally asymmetric with respect to the group paths that produced high digital leadership in manufacturing enterprise.

Based on the theory of planned behavior，the qualitative comparative analysis was used to explore the complex antecedent mechanism affecting the innovation willingness of scientific and technological talents in universities.The results show that internal ecological degree，external ecological degree，directive norm，exemplary norm，self-efficacy and control are not the necessary conditions for the innovation willingness of scientific and technological talents in universities； There are three paths to affect the innovation willingness of scientific and technological talents in universities，namely，innovation attitude-driven，directive-normative attitude-driven，and mixed-driven.The conclusions are helpful to systematically explain the complex antecedents of the innovation willingness of scientific and technological talents in universities，improve the innovation efficiency of scientific and technological talents，and empower the high-quality development of scientific and technological innovation.

In order to solve the thermo-mechanical fatgue damage problem of trubine blade under the complex multifield coupling service conditions，combining the improved Morrow low-cycle fatigue damage and creep damage models based on the load spectrum of the engine’s service conditions，the prediction of the thermo-mechanical fatgue life of engine turbine blade was achived.Firstly， considering the rotor-stator interference effect between cascades， the three-dimensional flow field modeling and multi-field coupling simulation of the turbine blade were completed. Further， based on the engine load spectrum and multi-field coupling response characteristics under service operating condition， the key assessment positions of fatigue damage of the turbine blade were determined. then， an improved Morrow low-cycle fatigue damage model was developed and compared with traditional models and experimental results for verification. Finally， using the linear cumulative damage criterion， as well as the Morrow low-cycle fatigue damage and L-M creep damage， thermo-mechanical fatigue life of the turbine blade was predicted under service conditions. The results show that the numerical simulation results are in good agreement with the experimental data， verifying the accuracy of the three-dimensional unsteady flow field simulation of the turbine blade and the improved Morrow low-cycle fatigue damage model. Under the engine load spectrum and multi-field coupling effect of service conditions， considering the Morrow low-cycle fatigue damage and L-M creep damage， the thermo-mechanical fatigue life of the turbine blade is 6.028×10³ h. The area with the minimum life is located at the leading edge of the blade root at the air inlet， the assessment position A， which is the key maintenance part of the turbine blade. This study can provide a theoretical reference and basis for the thermo-mechanical fatigue life assessment of turbine blade under complex service operating condition.

In order to study the impact of the Reynold number on the performance of wind tunnel heat exchangers，the CFD numerical simulation method was employed to analyze a conventional finned-type heat exchanger in a wind tunnel. Firstly，a three-dimensional geometric model of the finned-type heat exchanger was created through NX12.0，and then the mesh was generated through Ansys-Meshing. Numerical simulations were carried out using Fluent 2021R1. The numerical simulation mainly focused on the impact of the incoming flow Reynolds number on the heat transfer performance and resistance performance of the heat exchanger. The calculations revealed that corresponding to 2mm，4mm，and 6mm，the Reynolds numbers are 676.79，1 353.59，and 2 030.39 respectively. As the Reynolds number increases，the pressure drops and temperature difference between the inlet and outlet of the heat exchanger decreases，and the comprehensive heat transfer performance increases by 72.61% and 28.28% respectively. However，the improvement effect tends to be flat. Under identical structural parameters，as the inlet wind velocity increases，the corresponding Reynolds numbers are 1 355.09，2 710.18，4 065.27，5 420.354，and 6 775.44 respectively. As the Reynolds number increases，the heat transfer factor decreases，and the heat transfer characteristics of the wind tunnel heat exchanger show a downward trend. Meanwhile，the friction factor also decreases，leading to a downward trend in the flow resistance of the wind tunnel heat exchanger. The comprehensive heat transfer factor decreases，which decreases by 9.34%，8.96%，4.79%，and 5.34% respectively. Consequently，the comprehensive performance of the heat exchanger demonstrates progressive deterioration.

In order to investigate the drag reduction characteristics of groove structure，triangular and trapezoidal grooves with different dimensions were selected and simulation comparison on the drag reduction characteristics of flat models with two-dimensional transverse，three-dimensional transverse and three-dimensional longitudinal groove arrangements were carried out. The results show that the groove structure can generate low-speed fluid at the bottom of the groove. The low-speed fluid in the transverse groove can act as a rolling bearing. The drag reduction mechanism of the longitudinal groove can be explained from the perspective of protrusion height theory. The low-speed fluid in the grooves reduces the near-wall velocity gradient，thereby reducing friction drag. The groove structure can effectively reduce the turbulent kinetic energy and shear stress，leading to a reduction in viscous drag. The drag reduction effects of the two-dimensional transverse and the three-dimensional transverse groove model is similar，whereas longitudinal grooves exhibit a higher drag reduction rate than transverse grooves. Moreover，trapezoidal grooves achieve a higher drag reduction rate than triangular grooves. The optimal groove dimensions are a width of 0.1mm and a depth of 0.1mm，yielding a maximum drag reduction rate of 18.57%.

Addressing the high-precision modeling requirements of unsteady aerodynamics during complex aircraft maneuvers， a method for modeling non-steady aerodynamic forces based on an adaptive genetic algorithm （AGA） optimized long short-term memory （LSTM） neural network was proposed. Computational fluid dynamics （CFD） simulations were conducted to capture maneuver flight data during rapid turns at varying bank angles and rolling and looping maneuvers at different Mach numbers. An AGA-LSTM model was developed using this data to predict aerodynamic coefficients under non-steady conditions. Specifically， predictions for the aerodynamic coefficients during a 60° bank angle rapid turn maneuver were made， demonstrating accurate estimation of lift coefficient， drag coefficient， and pitch moment coefficient that closely matched CFD simulation results. To further validate the proposed model’s accuracy， predictions were compared with CFD simulation data and a traditional LSTM neural network model for Envelopment maneuvers. The results indicate that the AGA-LSTM neural network model provides closer predictions to simulation data compared to traditional LSTM models， thus offering improved prediction accuracy.

In order to prevent the issue of transmission efficiency decline or even failure in universal joint caused by wear，the wear amount was calculated based on the Archard model and Hertz contact theory.the functional relationship between the variation amplitude of the output angular velocity and the wear amount was given by using dynamic simulations，and a reliability and sensitivity calculation model was established.The results show that greater wear leads to greater fluctuation of the output angular velocity，with a significant increase in fluctuations once the wear amount exceeds a certain threshold.As the wear amount increases，reliability gradually decreases，and the sensitivity of reliability to each random variable is negative，with the parameter K _H having the highest sensitivity.Therefore，while regularly monitoring the wear amount and the variation amplitude of output angular velocity，appropriate process measures should be taken to improve material hardness and use high-quality lubricants to reduce the mean and dispersion of K _H.

In response to the need for rapid iteration in the conceptual design stage of propeller aircraft， low-fidelity aerodynamic evaluation methods， such as the vortex lattice method was a more suitable choice. In order to obtain suitable computation parameters and quantitative errors， the NASA OpenVSP unsteady vortex lattice method （UVLM） was studied using the propeller standard model， and the computation characteristics of the multiple reference frame quasi-steady vortex lattice method （MRF-VLM） were presented for the first time. The computational convergence and error characteristics were analyzed using the APC electric-propeller standard model， and the grid/iterative parameter settings that take into account both computational stability and efficiency were obtained. The comparison with experimental data shows that with appropriate grid and iterative parameters， the computation errors for the propulsion efficiency of the above two methods are within 6.1% and 3.6% respectively， under conditions of low pitch angle and medium advance ratio. The accuracy meets the requirements of the conceptual design stage， and the MRF-VLM computation with 4-thread parallel processing takes only 4 minutes， which is more efficient. The case of NACA 5868-9 propeller standard model further verify the computational reliability of the MRF-VLM method. The above research results can provide a reference for the reliable engineering application of OpenVSP VLM. However， due to the limitations of the linearized potential flow theory of the VLM， the above two methods can not accurately simulate the strong flow nonlinearity under high pitch angle and high/low advance ratio， and the computation accuracy needs to be improved.

To ensure the normal operation of turbine blades Facing continuously increasing turbine inlet temperatures， the development of more efficient cooling technologies has become particularly urgent. Ceramic matrix composite （CMC） transpiration cooling technology combines the excellent high-temperature resistance of CMC with the efficient heat dissipation potential of transpiration cooling， showing broad application prospects in future thermal protection of turbine blades. However， the strong anisotropy of CMC materials and their complex internal porous structure pose significant challenges for the thermal analysis of the transpiration cooling process， and the underlying flow and heat transfer mechanisms remain unclear. This review summarized recent research progress on the anisotropic thermal conductivity and microporous seepage characteristics of CMC， analyzed key existing problems and challenges in current studies， and offerd suggestions for establishing a comprehensive design framework for CMC turbine blade transpiration cooling structures.

Traditional transmission line wire crimping is done manually， and its crimping accuracy and consistency are difficult to ensure. To this end， an automatic crimping control system for transmission line conductors was designed， which took Siemens S7-1200 as the controller， took fuzzy PID as the core control algorithm， and applied the SCL language of Protherm platform for the programming and realization of fuzzy PID control algorithm. The designed system could set crimping parameters through the monitoring interface of the upper computer， realizing the functions of automatic movement of the slide table， automatic crimping of wires， automatic measurement of the distance between wires and edges， etc.， and the real-time change curves， the current variable values and the crimping status in the operation process of the system could be viewed in the monitoring interface. Experimental results show that the designed system can adaptively and dynamically adjust the control parameters to achieve precise control of wire crimping， improving crimping accuracy and reducing the labor costs.

To meet the requirements for open-loop trajectory motion of a solar-powered car， a design and fabrication method was proposed for a solar-powered car. Firstly， by using B-spline curves， the characteristic points were fitted to generate an ideal open-loop trajectory curve. MATLAB software was then utilized to analyze and optimize the planned trajectory to produce the cam profile. Next， based on the overall functional requirements， the overall structure， transmission mechanism， and steering fine-tuning mechanism of the solar-powered car were designed. Subsequently， simulation software was employed to perform simulation analysis and optimization of the cam trajectory， and the car’s circuit modules were assembled. Finally， actual driving tests of the car were conducted. To improve the trajectory accuracy of the car， the front wheel offset angle was modified by adjusting the fine-tuning mechanism. Experimental results demonstrate that the car fabricated using this method operates stably and successfully passes through all predetermined characteristic points.

To address the collaborative search task planning issue for unmanned aerial vehicles （UAV） and unmanned ground vehicles （UGV） in search and rescue scenarios， a method that combines centralized task allocation with independent path planning was proposed. Firstly， tasks were clarified based on requirements and assigned to UAV and UGV. Firstly， the UAV and UGV conducted trajectory planning and path planning independently according to their assigned tasks. For the task allocation problem，a multi-objective area segmentation and allocation model based on task types was proposed，and an adaptive hybrid algorithnm incorporating genetic algorithm and tabu search was developed in solution.For trajectory planning， a “traveling salesman path-area coverage path planning” model was constructed based on the task requirements of the UAV， employing an improved genetic algorithm that introduced an “S”-shaped path as the initial population and designed a fitness function to optimize the selection process. Regarding the path planning for UGV， a method that combines A* algorithm with artificial potential field algorithm was proposed to find the optimal traveling path. Simulation results indicate that the proposed algorithm effectively completes search tasks and shows significant improvements in both task execution efficiency and path planning accuracy compared to commonly used genetic algorithms and A* algorithms.

As one of the critical components of aircraft engine icing，the water droplet impingement characteristics on the rotating fairing surface directly influences the subsequent icing state.To investigate the water droplet impingement behavior under different operating conditions，a three-dimensional water droplet impingement model for the rotating fairing of aero-engine was established by using the Euler method and single rotating coordinate system.The water droplet impingement characteristics of the rotating fairing under stationary and rotating conditions were simulated respectively.The results show that under stationary conditions，as the freestream Mach number increases，both the water droplet collection coefficient and impingement area​ exhibit significant growth； compared to single-diameter droplets，when considering the Langmuir-D distribution​​of water droplets，the collection coefficient at the stagnation point decreases，while the impingement zone increases and the downstream collection coefficient also slightly increases.Under rotating conditions，the​​rotational speed​​ has negligible effects on droplet impingement characteristics due to the streamlined aerodynamic profile of the rotating fairing.

Tibetan Jiu Chess， a traditional folk chess game， is a complete information game that carries the profound Tibetan civilization and splendid culture. In view of the complexity of the rule system and the diversity of the game changes， the traditional game search algorithm is unable to cope with the vast game board and complex strategies. In order to improve the intelligence level of Tibetan Jiu Chess， a Monte Carlo tree search （MCTS） algorithm optimization strategy incorporating prior knowledge was proposed. The strategy was based on deep reinforcement learning in the key phases of layout planning and move strategy，and the strategy selection optimization function and evaluation function were designed by integrating the prior knowledge of domain experts. The search process of MCTS was efficiently guided by functions，and the best model for high-quality tessellation could be trained. Experimental results show that the improved MCTS algorithm achieves significant performance in the game.

Nomex honeycomb materials are widely used in marine，high-speed rail，aerospace and other fields because of the material’s low density，high specific strength and high specific stiffness. However，the non-homogeneous，thin-walled and brittle characteristics of honeycomb structures make them prone to tearing，crushing and other defects during processing. Therefore，scholars both in China and abroad have used ultrasonic machining technology to solve the problem of Nomex honeycomb material processing with the advantages of small cutting force and high surface quality，which have carried out extensive and in-depth research. However，compared with ordinary machining，high-frequency vibration and periodic impact generated in high-quality and efficient ultrasonic machining process put forward higher requirements for tool performance and life. Therefore， focused on tool material selection， tool design， kinematic analysis， cutting force and cutting heat modeling， and surface topography analysis during machining of Nomex honeycomb materials. The research status of machining mechanism，cutting characteristics and ultrasonic tool design for ultrasonic cutting of Nomex honeycomb materials were summarized. The future research and development trend of ultrasonic cutting Nomex honeycomb materials were put forward， which provided a reference for further research on the service performance and processing quality of ultrasonic cutting tools in cutting.

In order to better solve the problem of robust fault-tolerant control for magnetorheological landing gear buffer systems with damper failures， a model reference adaptive fault-tolerant control strategy was proposed，and a mechanical model was established and key parameters were determined by damping experiments. A fault model was established for the magnetorheological landing gear system and fault-tolerant control was introduced. An adaptive law was designed to adjust the control gain in real time， and a fault-tolerant controller was constructed to deal with damper failures. Results show that compared with the traditional passive and model reference adaptive control methods， the buffer efficiency of the damper fault tolerant control method is improved by 4.9% and 0.95% respectively. This control method significantly improves the damping efficiency and quality of the magnetorheological landing gear damping system.

In order to solve the problem that the current fitness exercise evaluation systems provided poor evaluation information and involved high computational complexity， a fitness exercise evaluation system based on human posture estimation was proposed. The system utilized the built-in camera of mobile phones or tablet computers， combined with human pose estimation algorithms， action recognition algorithms， and action evaluation algorithms to achieve intelligent and effective evaluation of users’ fitness exercises. Firstly， the MediaPipe algorithm was used to estimate the human posture and obtain the human joint bone data.Then， it was fed into the proposed SCBT-GCN network for action recognition. Based on kinematic principles， a series of evaluation methods had been designed which evaluate joint angles， symmetry， trajectory characteristics， and movement fluency. Finally， corresponding targeted movement assessment algorithms were invoked based on the identified action categories to detect and evaluate abnormal fitness movements， thereby forming a fitness action evaluation system with strong intelligence and high real-time performance. Experimental results show that the system has an accuracy of 98.6% in action recognition， which can detect abnormal actions in real time and score fitness actions offline， demonstrating high practical value.

To ensure the dynamic performance of the electric helicopter equipment platform under complex vibration and shock environments， a combined test and simulation approach was adopted to conduct an in-depth investigation of a specific electric helicopter equipment platform. The simulation part utilized HYPERMESH and ANSYS software to perform modal analysis， frequency sweep analysis， random vibration analysis， and transient response analysis， obtaining the platform’s dynamic characteristics and key data. The test part included random vibration and shock tests. The random vibration testing employed an acceleration power spectral density （PSD） scaled to 1g RMS as the excitation condition， and the shock testing was conducted with a peak acceleration of 4 g and a half-sine wave excitation lasting 6 ms. The results show that the first two modal frequencies of the platform are 33.735 Hz and 39.751 Hz， and the low-frequency modes may couple with low-frequency excitations during operating condition. Random vibration analysis shows that the primary response is concentrated within the 70~300 Hz range. The maximum deformation of the platform under random vibration conditions is 1.572 9 mm， and the stress distribution is uniform， meeting the material yield strength requirements. Transient shock analysis indicates that the platform’s maximum stress under a 4 g shock is 11.464 MPa， which is significantly lower than the material yield strength. The correlation between test and simulation results verifies the reliability of the platform design. This study provides a theoretical basis and reference for the optimized design of electric helicopter equipment platforms.

In the context of cooperative operation between UAV and ground vehicle， rotor UAV often needs to pass through necessary points or avoid obstacles when executing related tasks， which increases the trajectory fluctuation and brings more challenges to path planning. To solve this problem， an improved discrete mechanics and optimal control （DMOC） trajectory optimization algorithm was proposed， which transformed the optimal control problem into a nonlinear programming problem. Considering the endurance time of vehicle-mounted rotor UAV， the shortest time was set as the optimization objective under guaranteed traversal conditions. In the process of the algorithm implementation， an approach was proposed to adjust the distance walk length according to the size of the error factor， and then approximate correlation integral， which effectively solved the problem of trajectory fluctuation. The experimental results show that this method can improve the smoothness of the optimal trajectory while ensuring that all the necessary points are traversed，and also provides a reference for the trajectory optimization of other agents.

In order to select the characteristic indicator to more effectively characterize the performance degradation of the air compressor bearing，a feature indicator selection method based on game theory combinatorial weighting method was proposed.Through the analysis of the time-domain indicator of the bearing vibration and the preprocessing of redundant information，the time-domain indicator dataset of the bearing was obtained.Three characteristic indicator evaluation methods including monotonicity，robustness and trend were used to complete the selection of performance characteristic indicators.Based on this，the game theory combinatorial weighting method was used to weight three characteristic indicator and sixteen time-domain indicator to complete the selection of bearing performance degradation indicator.The effectiveness of this method was illustrated by example verification.

In order to deeply study the driving performance of ionic liquid gel （ILG）as ionic electroactive polymer，developed a novel soft actuator based on ILG and conducted a detailed investigation into the electromechanical coupling model of the ILG soft actuator. According to the material properties and current response law of electroactive polymers，established electromechanical coupling equations and driving equations for ILG soft actuator based on the equivalent transformer model of ionic polymer-metal composite actuators proposed by Claudia Bonomo. The least squares method was used to identify the coupling model of the ILG soft actuator， and the influence of structural parameters on the end displacement and driving force of the soft actuator was analyzed， providing a theoretical basis for the control of soft actuators. The electromechanical coupling model of the ionic liquid gel soft actuator is established， which lays a foundation for the development of ionic liquid gel soft robots.

To investigate the influence of assembly deviations on pipeline sealing performance， a finite element model was established based on the torque-tension relationship. The influence of three assembly deviations （axial，radial， and angular） acting individually on the sealing performance of pipeline connectors was analyzed， and the coupling influence of these three assembly deviations on the sealing performance of pipeline connectors was studied through orthogonal experiment. The results indicate that under single deviation conditions， negative axial deviation，radial deviation，and angular deviationall reduce pipeline sealing performance. When the three assembly deviations are coupled， the sealing performance decreases. Negative axial deviation and angular deviation have significant influences on pipeline sealing performance， while the influence of radial deviation is relatively small. Therefore， in practical engineering， negative axial deviation and angular deviation should be strictly controlled， and positive axial deviation and radial deviation should be minimized as much as possible.

Aiming to address the limitations of traditional health monitoring methods of hydraulic oil pump truck testbed， such as reliance on periodic physical inspections and maintenance， low levels of informatization， and insufficient production efficiency， an application development architecture for a hydraulic oil pump truck monitoring system was proposed based on digital twin technology. This architecture integrates data， models， and communication to improved the predictive maintenance level and operational efficiency of the equipment. The research enabled real-time data interaction and fusion between the physical equipment and the virtual model by constructing a digital twin virtual model of the equipment and utilizing the TCP/IP protocol and a unified JSON packet format. On this basis， the system could perform simulation， numerical prediction， fault diagnosis， and early warning for the hydraulic oil pump truck testbed， thereby optimizing the equipment’s operation status.Test results show that the system is able to perform health monitoring and fault diagnosis for the hydraulic oil pump truck， showing broad prospects in the application of aviation testbed.

The key to achieving high level of scientific and technological self-reliance is to strengthen the incentives for the innovative behaviors of scientific and technological personnel. Based on the AMO theory， combined with NCA and QCA methods， a driving path of scientific and technological personnel’s innovative behaviors from the perspective of complex interactive configuration between individuals and organizations was explored. The results show that the emergence of a single factor is not a necessary condition for the outcome variable. There are three paths that can drive the high innovative behavior of scientific and technological personnel. It enriches the relevant theoretical research on the scientific and technological personnel’s innovative behavior， providing reliable basis and concrete measures for managers to effectively drive the innovative behavior of scientific and technological personnel.

To enhance the resilience of drone enterprise supply chains and their ability to respond to risks， the impact and mechanism of artificial intelligence technology on supply chain resilience were empirically examined which based on data from A-share listed drone companies from 2016 to approximately 2022. It was found that strengthening artificial intelligence technology could significantly improve supply chain resilience. This conclusion remained valid after a series of robustness tests and was particularly pronounced for small and medium-sized drone companies with lower market share. Mechanism analysis indicated that artificial intelligence mainly operates through two pathways： enhancing supply chain concentration and reducing internal management costs. By optimizing resource allocation， improving operational efficiency， and reducing management expenses， it strengthened the enterprise's ability to cope with market fluctuations and external shocks. This study provided policy and strategic recommendations for governments and companied to use artificial intelligence to optimize supply chain management， as well as the oretical support and practical guidance for enhancing enterprise risk resilience.

To efficiently predict the dynamic stability of integrated thermal protection systems under aerodynamic heating environment， a thermal flutter calculation method for corrugated sandwich panels based on a thermo-mechanical equivalent model was proposed， aiming to enhance the computational efficiency of finite element analysis. First， the homogenization theory was applied to approximate the core layer as a single layer of orthotropic material， thereby simplifying the model. Second， the proportion of the areas occupied by the web and the insulating material was redefined to increase the characteristic size of the equivalent area， thereby reducing the number of finite element grids while ensuring the accuracy of the temperature gradient. Finally， based on the developed thermo-mechanical equivalent model， the modal characteristics of the corrugated sandwich panel were extracted， and the p-k method was employed for thermal flutter analysis to evaluate the aerodynamic response. The results show that this method improves computational efficiency by 80%， and the error is controlled within 3%.

In the highly complex and intensely adversarial air combat environment， current unmanned aerial vehicle intention assessment methods generally face challenges such as strong subjectivity of fusion rules， ignoring the time correlation of relevant attributes， and insufficient defect information processing means. To address these issues， this paper proposes a dynamic evidence network method that integrates defect information correction and subjective and objective rules. Firstly， focusing on the strong correlation of continuous variables in the time dimension， a spatiotemporal fusion evidence network model was modularly constructed. Subsequently， by introducing the LSTM trajectory prediction technology， a defect evidence correction mechanism was established， which significantly improved the accuracy and integrity of information. Finally， objective rules were designed based on statistical calculation methods， and combined with subjective experience， a library of subjective and objective rules was constructed. Based on the above improvements， simulation experiments were conducted. The results confirm that the proposed mechanism of defect information correction， subjective and objective rules， and dynamic fusion have significant advantages in improving the accuracy and credibility of intention recognition results.

To verify whether the mechanical properties of the composite skin of a certain type of electric aircraft meet the design requirements under hot and humid environments， an environmental chamber was used to simulate the actual operating conditions， and the conditioning of test specimens under elevated-temperature and wet conditions was completed. for composite foam sandwich structure wing skin instability test specimens from a certain type of aircraft， the environmental chamber simulation technology was adopted. The elevated temperature wet （ETW） conditioning was completed during the preliminary assessment of test conditions. Comparative analysis of failure modes and structural strength was conducted under both room temperature dry （RTD） and ETW environments. The strain variation with compressive load was monitored， and the influence of the hygrothermal environment on the mechanical properties of composite materials was analyzed. The results indicate that the recommended maximum service temperature for the structure of this test specimen is 71 ℃， with a minimum temperature of -54 °C and a relative humidity of 85% RH. Post-test analysis reveals that although both the strength and stiffness of the structure are significantly reduced under ETW environments， they still meet the airworthiness verification requirements. Due to the influence of the ETW environment， the collaborative working ability between the fibers and the matrix is weakened， and they can’t efficiently cooperate to jointly bear the load and resist deformation. Therefore， the failure modes of the test specimen under the ETW environment exhibit greater complexity and diversity compared to those observed under the RTD environment. The results provide data support for optimizing the environmental conditioning scheme， establishing environmental design criteria， and completing the airworthiness verification plan， while offering critical technical references for designing composite structure in aircraft operating under complex environmental conditions.

To accurately capture the aerodynamic characteristics of Z-shaped folding wing at different folding angles， while considering the influence of the structural elastic effect of different wing segments， a parametric aerodynamic-structural coupling model was constructed based on a quasi-steady state environment.Using ANSYS Workbench software with the RNG k - \epsilon turbulence model and the Coupled algorithm，numerical simulation was conducted to analyze the aerodynamic characteristics of the folding wing under the action of elastic effect and the results were compared with those of a rigid wing. Additionally， the deformation laws of the elastic folding wing in different flight environments were systematically explored. The results show that under the influence of elastic effect， the absolute values of lift and pitching moment of the folding wing are slightly lower than those of the rigid wing. As the folding angle increases， the impact of elastic deformation on the overall aerodynamic performance of the wing stabilizes. When the folding angle is 90°， the differences in lift and pitching moment between the elastic and rigid wings are only 0.93% and 0.70% respectively. Furthermore， under the influence of elastic effect， the maximum deformation of the wing occurs at the wingtip. As the folding angle increases， the magnitude of wingtip deformation gradually decreases.

In order to improve the accuracy of component performance evaluation under a whole-engine test environment for aero-engine，a performance optimization and evaluation method was proposed for aero-engine components based on the artificial bee colony algorithm. This method took the performance parameters of the components to be evaluated as input variables. It utilized an overall aero-engine performance simulation model combined with the ABC algorithm to optimize and obtain the optimal set of performance parameters for the components to be evaluated， ensuring compliance with the accuracy requirements of the test parameters. This method was applied to evaluate the component performance under the test environment of a core engine. The results show that the deviation between the evaluated parameters and the measured values remains within 0.95%， demonstrating the high effectiveness and engineering practicability of the evaluation method. During the multi-operating condition optimization process， to balance the accuracy of the entire speed range， the maximum deviation increases by approximately 0.1%. But the introduction of solution constraints enhances the authenticity of the performance evaluation and reduces the impact of test uncertainties on the performance evaluations. Meanwhile， the correction characteristics of component performance under different inlet temperature environments of the core engine are obtained in the optimization process.

To improve the calculation accuracy of the overall performance model， a overall performance modeling method for aero-engines with primary and secondary flow integration was investigated. Through research on principles for aero-engine performance calculation， the influencing factors of secondary flow on the primary flow thermodynamic cycle were comprehensively identified. Their mechanisms of action were clarified. A characteristic model of the impact of secondary flow on the primary flow thermodynamic cycle was constructed. Based on the traditional modeling framework， this method innovatively introduced a coupling calculation mechanism for secondary flow rate， energy， and mixing terms with the primary flow thermodynamic cycle. Using this method， an overall performance calculation model was developed for a specific aero-engine. The results show that， compared to traditional models， the proposed method achieves a 1.4% improvement in thrust calculation accuracy， a 1.33% enhancement in fuel consumption rate prediction， and a 2.7% increase in turbine inlet temperature estimation accuracy， significantly optimizing the precision of the overall performance calculation model for aero-engines.

To improve visual tracking precision for friction stir welding under weld seam occlusion and other interference， a weld seam feature tracking algorithm based on the ECO-HC was proposed. A traditional image processing technology was employed to detect the initial weld seam feature point， while enhancing the sensitivity of the algorithm to abnormal disturbances by introducing a dual confidence assessment mechanism comprising similarity calculation and peak to sidelobe ratio on the basis of the ECO-HC algorithm. Additionally， a trajectory prediction method based on curve fitting was proposed to achieve the relocalization of lost target. Experiments were conducted on aluminum alloy weldments of varying thicknesses. The experimental results show that the mean absolute error of the weld seam tracking system proposed can be maintained within 0.051 mm， which fully meets the precision requirement for weld seam tracking and demonstrates the effectiveness of the algorithm proposed.

In order to solve the problems of large volume， limited travel and aseptic isolation requirements of vascular interventional surgery robot， a slave mechanism of vascular interventional surgery robot imitating doctor’s action was designed. The bionic finger-based mechanism enabled unrestricted continuous delivery and rotation of guidewires / microcatheters， simultaneous rotation and delivery of guidewires， and adjustable clamping distance and force and the modular design was utilized to achieve the isolation of sterile/non-sterile operation areas.A three-dimensional model of the mechanism was developed， and the design structure was analyzed kinematically to establish its kinematic equations.The rationality of the design was validated through simulation analysis. Experimental studies were conducted to investigate the positioning accuracy of guidewire delivery. Results indicate that the mechanism achieves a positioning error within from -0.5 mm to 0.5 mm under static conditions， enabling precise delivery of guidewires. This performance meets the clinical requirements for vascular interventional surgery robot.

In order to study the influence law of main process parameters on forming limit of conical parts in dieless spinning based on spherical roller， the influence of four main process parameters， such as spinning part half-cone angle， feed rate of the spinning roller， initial diameter of the sheet metal and roller radius， on the forming limit of the spinning part was studied through the spinning forming experiment of the conical part with spherical roller. The results show that the decrease of the half-cone angle of the spinning part leads to the increase of the feed of the spinning roller in the axial direction of the sheet material and the increase of the contact area between the sheet material and the spinning part roller. The spinning part is prone to wrinkles or cracking. In addition， the decrease of the half-cone angle of the spinning part will increase the bending degree of the sheet material at the round corner of the mandrel， and the spinning part is prone to crack. The contact area between the sheet and the roller increases with increasing the feed rate of the roller. The spinning part is prone to appear wrinkle. When the feed rate of the roller decreases， the forming time of the roller to the sheet is prolonged， and the spinning part is prone to appear crack. The increase of the initial diameter of the sheet metal increases the time of the tensile stress on the radial direction of the sheet metal， which leads to the cracking of the spinning part at the top fillet. While the decrease of the half-cone angle of the spinning part leads to the increase of the deformation of the sheet metal and the risk of wrinkling of the part flange. The flange of the part is more likely to wrinkle. The radial tensile stress of the sheet metal with a smaller radius is longer in duration during forming， which leads to cracking of spinning parts more easily. The circumferential compressive stress of the sheet metal with a larger radius increases and lasts longer in duration at the end of spinning forming， resulting in the outer edges of the spinning parts being prone to wrinkles.

Automatic drilling technology plays an important role in modern aircraft manufacturing. To improve the efficiency and reduce the cost of aircraft assembly and manufacturing，a hole path planning algorithm of aircraft curved surface components was studied in depth. Based on the situation that the traditional ant colony algorithm was easy to fall into local optimum and converge slowly，the new school-based optimization algorithm combined with the traditional ant colony optimization algorithm was proposed to achieve the optimization of the drilling path of aircraft curved surface components. The simulation results show that the improved school-based optimization ant colony algorithm is superior to the traditional ant colony algorithms in terms of drilling path length and iteration speed.

Simulation and experimental comparison methods were used to study the thermal expansion coefficient and anisotropic thermal expansion behavior of 2.5D SiC_f /SiC ceramic matrix composites in different directions to provide certain ideas for the thermodynamic research. Thermal expansion coefficient of 12 temperature gradients in the range of 100~1 200 ℃ in the 0°，45° and 90° directions of 2.5D SiC_f /SiC ceramic matrix composite were measured respectively to analyze the influence of direction and temperature. Then simulation models in different directions were established to simulate the experimental heating process and to get thermal expansion coefficient of the material. The results show that the coefficient of thermal expansion increases with temperature rises and then gradually stabilizes. In terms of direction， the thermal expansion coefficient of the material in the direction of 45° is greater than that of the other two directions， and in the direction of 90° is the smallest. The thermal expansion coefficient is calculated through the thermal deformation data results of the simulation model. The calculated Pearson correlation coefficients are 0.968 06， 0.974 19 and 0.917 37， which proves the validity of the model and lays the foundation for future research.

To address the issue of poor system adaptability inherent in traditional event-triggered mechanisms，a problem of dynamic event-triggered consensus of uncertain multi-agent systems under switching topology was studied. The event-triggered mechanism could not effectively guarantee the system adaptability in practice. It was proposed to use the dynamic event-triggered mechanism to dynamically adjust the current state of the system， which could reduce the frequency of the system update. In designing the consensus controller， factors such as switching topology， time delays， and external disturbances were considered. Algebraic graph theory and related theoretical methods proved that the dynamic event-triggered mechanism set the consensus controller to achieve the consensus of the uncertain multi-agent system. Finally， the simulation results confirmed that the method is reasonable.

To solve the problem of high false detection rates in tunnel fire detection caused by the complexity of tunnel environments based on the YOLOv8n network model， an improved tunnel fire detection algorithm was proposed.First， in the backbone network， the FasterNet network was used for replacement while retaining the original SPPF module to achieve more comprehensive feature extraction； Secondly， in order to improve the detection accuracy of the model for irregular targets in the complex background， the D-LKA attention mechanism was introduced in the C2f module； Finally， Focaler-IoU to optimize the model loss function was introduced， which further reducing the problem of false positives or false negatives caused by distractors. The experimental results show that compared with YOLOv5， YOLOv7 and the original models of YOLOv8n， the accuracy of the improved model is increased by 7.6%， 5.6%， and 3.5% respectively， and the average accuracy means are increased by 8.3%， 7.7%， and 5.1% respectively. Compared with other YOLOv8n-based improved algorithms， the mean average precision of our proposed model is increased by 3.3% and 6.4% respectively.

To gain an in-depth understanding of the latest research advances in gait recognition， the existing studies were comprehensively reviewed and summarized， with a focus on discussing the strengths and weaknesses of deep learning-based gait recognition methods， as well as outlining their applications and prospects in civil aviation security. The results show that deep learning methods have significantly improved the accuracy and robustness of gait recognition， demonstrating greater potential in civil aviation security scenarios such as suspectassisted identification and intelligent earlywarning.