In order to address the challenges in structural strength analysis caused by limited training samples and extreme environmental conditions while improving analytical efficiency，deep transfer learning methods was applied to investigate the mechanical properties of composite laminates for the RX4E electric aircraft.Based on the analysis of experimental results obtained from composite laminates， multiple deep learning models were compared in terms of their ability to predict the experimental results. Finally， the convolutional long short-term memory （CLSTM） was selected as the optimal deep learning model. Furthermore， a transfer learning （TL） model was introduced to accurately predict the stress-strain relationships of composite laminates under varying temperatures， humidities and layup configurations. The results indicate that the proposed TL-CLSTM network model has exceptional capability in predicting the mechanical properties of composites， particularly in predicting the stress-strain relationship， with a mean squared error and a root-mean-square error of 10-5and 10-3 respectively.The proposed model can effectively predict the mechanical properties of composite laminates for electric aircraft， overcoming the complexities and inefficiencies of traditional mechanical properties measurement methods，which providing a novel pathway for the future study of electric aircraft manufacturing.
In view of the problem of traditional continuous carbon fiber toughened ultra-high temperature ceramic matrix composites （Cf /UHTCs）， such as long preparation period and high cost， an internal grouting process of carbon fiber based on ultrasonic regulation technology was proposed to realize the uniform distribution of ceramic powder in carbon fiber braided body， and the mechanical properties of the prepared composites were analyzed. The research results show that the mechanical properties of the composites can be effectively improved by using ultrasonic regulation technology to assist the internal grouting of carbon fiber， and it has a remarkable effect in the preparation of ultra-high temperature ceramic matrix composites toughened by continuous carbon fiber， which can improve the material preparation efficiency and improve the material properties at the same time， and plays a very important role in promoting the innovation of composite preparation technology.
To reduce the development costs of electric aircraft， a multidisciplinary optimization design approach was investigated based on the entire life cycle. By analyzing the composition of the entire life cycle costs for electric aircraft，an entire life cycle cost estimation model was developed. The research established multidisciplinary analysis models encompassing aerodynamics， flight performance，mass， and economic factors. With the optimization objective of minimizing life entire cycle costs， a comprehensive multidisciplinary optimization design method for electric aircraft was proposed. Taking a four-seat electric aircraft as a case， the parallel subspace optimization algorithm was employed to identify the optimal solution， resulting in an optimized overall design scheme that validates the method's effectiveness. This method can be applied to electric aircraft conceptual design， providing technical support for electric aircraft development.
To address the control requirements for extending the residual stability margin of adaptive variable-cycle engines， a specific configuration of an adaptive variable-cycle engine was selected as the research object. The basic effect sensitivity analysis method was used to analyze the impact of geometric adjustment on the stability margin under three typical engine operating conditions： ground take-off state， subsonic cruise state and supersonic cruise state. The adjustable mechanisms that should be prioritized when the stability margin of the compression system of the adaptive variable-cycle engine was insufficient were identified. The geometric adjustment stability margin extension law was established for this specific adaptive variable-cycle engine and verified under typical operating conditions. The results show that adjusting the throat area of the outer nozzle is effective for extending the front fan stability margin； adjusting the throat area of the nozzle and the ejector area of the rear bypass is effective for extending the rear fan stability margin； adjusting the compressor inlet guide vane angle is effective for extending the compressor stability margin. When applying the established geometric adjustment strategy under typical conditions， the absolute error in stability margins for all compression components is less than 0.5%， demonstrating the good applicability of the proposed method.
In order to improve the hot corrosion resistance of TiAl-based alloys in molten salt， rare earth Ce modified NiCrAlY coating was prepared on the surface of TiAl-based alloy by atmospheric plasma spraying technology. The hot corrosion tests in molten salt of NiCrAlY（Ce） coatings were carried out at 900 ℃ for 60 h. The effect of rare earth element Ce on the hot corrosion behavior and mechanism was further discussed. The results show that the addition of rare earth element Ce can improve the hot corrosion resistance of NiCrAlY coating. After hot corrosion at 900 ℃ for 60 h， the corrosion layer on the surface of the NiCrAlY/TiAl coating system peels off severely， causing the failure of the coating. Therefore， the TiAl alloy has suffered from severe corrosion. For the NiCrAlYCe/TiAl coating system， its surface still covers with a complete corrosion layer and only small cracks generat at the interface between the coating and substrate. The NiCrAlYCe coating still possesses hot corrosion resistance property to a certain degree after corrosion.The addition of rare earth element Ce is beneficial to the selective oxidation of Al element， which hindered the formation of spinel oxides. The dense Al2O3 film blocks the internal diffusion of S element， thereby improving the corrosion resistance of NiCrAlY coating and extending service life of TiAl based alloy.
Aiming at the problem that the forming quality of wingtip fairing and glass fiber rib parts in aircraft body couldn’t be guaranteed in the manufacturing process， the influence of uniformity of temperature field distribution on mechanical properties and solidification deformation of parts was studied by optimizing the curing program， changing the number of parts entering the tank， and monitoring the surface temperature field distribution of parts by thermocouple.The results show that in the curing process， appropriately increasing the curing temperature can effectively increase the heating rate in the lower temperature area， improve the mechanical properties of parts and reduce the deformation during curing. When six glass fiber rib specimens are put into the autoclave， the heating rate in the whole tank will decrease， which will affect the curing effect of the structure. In actual production， the number of specimens in the autoclave should be reasonably arranged to ensure the forming quality.
Video person re-identification is a technology for identifying specific person in a multi-camera surveillance network. Compared to the methods based on single-frame images， this type of algorithms can provide more person information， but it also has issues such as model complexity and misalignment in constructing features. To address those issues， a feature fusion-based video person re-identification algorithm was proposed. The proposed algorithm included a global branch and a local branch with spatial transformation. The global branch extracted the global features of person， capturing coarse-grained information and overall contextual information of the person. The local branch with spatial transformation integrated a spatial transformation matrix into the local branch to learn discriminative local regional features and alleviating the issue of feature misalignment. By utilizing a multi-branch structure， the algorithm fused local and global features and aggregated features through temporal average pooling to enhance the diversity of features and improve the robustness of the model. Finally， the model was trained using cross-entropy and a soft boundary triplet loss. The test results on the Mars and DukeMTMC-Video datasets have verified the feasibility of the proposed algorithm. Specifically， the Mars dataset achieves mAP and Rank-1 accuracies of 82.25% and 89.76% respectively， demonstrating excellent practicality.
Regular inspection of transmission lines is an important guarantee for the long-term stable operation.To solve the problems existing in traditional manual inspection methods such as high labor intensity，low work efficiency and high cost，a PLC-based transmission line automatic inspection robot system was designed， focusing on the robot inspection operations of anti-vibration hammer resetting and broken lines repairing control problems. The manual and automatic control mode were used， the classic PID control algorithm was applied to reset the anti-vibration hammer， and repair the broken lines. The experimental results show that the transmission line inspection robot repair control system can accurately achieve the resetting of the anti-vibration hammer and the repairing of broken lines， and get a better control effect. The system designed is of great significance for the inspection robot to replace the traditional manual inspection method， reduces the labor intensity of inspectors and improves the automation level of transmission line management and maintenance.
To address the process route planning problem characterized by dynamic process requirements and intricate process data，a flexible process route planning method was proposed based on deep recurrent q-network （DRQN）.Firstly， leveraging the structural advantages of the long short-term memory （LSTM） network， sequential data features were thoroughly mined to enhance the accuracy and stability of process route planning.Secondly， by integrating the robust dynamic decision-making capability of the deep q-network （DQN） with an adaptive adjustment strategy， the challenges posed by fluctuations in requirements and processing environments were effectively mitigated.Lastly， in response to frequent process changes， a "selective forgetting" mechanism was implemented to improve the response speed of process route planning during step process changes.Simulation results demonstrate that the proposed method can efficiently resolve the process route planning issue associated with part occurrence feature reconstruction.
In order to improve the spoofing interference detection capability of satellite navigation system， a satellite navigation spoofing interference detection algorithm based on RNN was investigated， and the loss function was designed. In order to improve the accuracy of data prediction， a data preprocessing method was studied， which maped the data in a fixed interval and amplifies the characteristics of the data. The experimental results show that the prediction accuracy of the RNN model for the signal-to-noise ratio of ten satellites is higher than that of the Transformer model. The recurrent neural network model has an average accuracy of 64.76% in predicting the signal-to-noise ratio data， while the Transformer model has only 3%. In the RNN prediction model， the accuracy of the prediction for 7 out of 10 satellites signal-to-noise ratios is above 60%. It can be seen that the RNN model has a better prediction effect when facing the signal-to-noise ratio data of BeiDou satellite navigation signals with the time series data type. Therefore， the RNN model can realize the prediction of 0.08dB error for BeiDou signal-to-noise ratio， and when the difference between the future signal-to-noise ratio value and the predicted value is greater than 0.08 dB， it is considered that the signal is a spoofing signal at this time， so as to realize spoofing interference detection. The research results provide certain reference value for the research of satellite navigation spoofing algorithm.
In response to the current problems of relying on manual experience and slow processing speed for ligh electric helicopter faults，a fault diagnosis expert system based on fault tree was proposed to ensure the flight safety and reliability.Fault tree model was constructed based on common faults of a light electric helicopter. Through qualitative and quantitative analysis of the fault tree，the probability importance FVI values of each event in the fault tree were calculated. The FVI values were used to rank each fault event and convert the fault tree into a binary fault tree. An expert knowledge base was established through binary fault tree analysis，and the inference machine was designed based on the production rules and forward reasoning strategies in the knowledge base. Finally，the expert system was developed using QT，and the simulation testing proved that the designed expert system can effectively diagnose faults in light electric helicopters.
Based on the measurement of EEG，the behavioral intention of online shoppers after interacting with shopping website was evaluated and predicted.The research method combining objective electroencephalogram （EEG） experiments and subjective evaluation scales was employed to measure the EEG of online shoppers during the interaction tasks on the homepage of shopping websites， and to predict the behavioral intentions of online shoppers towards shopping websites after completing the tasks. The method of variance analysis was adopted to extract the significant EEG-based indicators which influence users’ behavior intentions. Based on the partial least squares （PLS）method，a relationship model between the users’ electrical indicators and the behavior intentions was established， and the model was verified. The results show that the proposed model can predict the approach trend of online shoppers towards shopping websites well.
