During the production of monocrystalline silicon, defects generated during the crystal pulling process are recognized to severely impact product quality. Traditional visual-based defect detection methods, when applied to the detection of small protrusions in crystal pulling images, are confronted with challenges such as slow detection speeds, large parameter volumes, and difficulties in deployment on embedded terminals. In response to these challenges, an improved YOLOv8 object detection model was proposed incorporating a ContextGuided module to enhance the inference efficiency of the model. An efficient DySample was introduced into the feature fusion network to optimize the efficiency and depth of feature fusion. A lightweight network structure was adopted to reduce the complexity and computational demands of the model, making it suitable for devices with limited computing resources. The model has been trained and tested on an industrial dataset, demonstrating a more accurate detection of small protrusions with a mean average precision (mAP) of 97.7%. Compared to YOLOv8n, it exhibits an increase of 11.6% in precision and a reduction in parameter volume by 31.9%, facilitating its deployment on embedded terminals.

In order to improve the output performance of permanent magnet assisted synchronous reluctance motor (PMa-SynRM), a multi-objective optimization design method for external rotor PMa-SynRM based on kernel extreme learning machine (KELM) and fast non-dominated sorting genetic algorithm (NSGA-II) was proposed. Firstly, the preliminary design of the PMa-SynRM rotor magnetic barrier was carried out and the working principle of the PMa-SynRM was analyzed. Secondly, the influence of each design variable on the optimization goal was evaluated through comprehensive sensitivity analysis, and the main optimization parameters were selected. Thirdly, with high output torque, high efficiency and low torque ripple as the optimization goals, a surrogate model based on KELM was established. Finally, NSGA-II was used for global optimization, and the optimal solution was selected from the Pareto frontier generated by NSGA-II, which was verified by finite element analysis. The simulation results show that the average torque of the optimized motor is increased by 15.83%, the torque ripple is reduced by 60.27%, and the efficiency of the optimized motor is also improved compared with the initial motor, which verifies the effectiveness of the optimized design method proposed in this paper.

Under the background that the accuracy of voltage amplitude detection is increasing year by year, a voltage amplitude detection method based on improved second-order generalized integrator was proposed to solve the problem that the existing voltage amplitude detection methods were susceptible to the interference of harmonic and DC components, which leaded to insufficient detection accuracy. First, the composition change of voltage after fault was analyzed, and the existence of harmonic component and DC component in fault voltage was proved. Then, the principle and performance of voltage amplitude detection by sine-cosine component method and traditional second-order generalized integral method were analyzed. Then, the output of traditional second-order generalized integrator without DC component was taken as the input of the improved second-order generalized integrator’s second module. In this way, the effective filtering of the DC component was realized, and the Bode diagram, zero pole and step response of the improved second-order generalized integrator under different gain coefficients were analyzed. The optimal gain coefficient was selected considering the influence of filtering effect, stability and response speed. Finally, the simulation results show that the method proposed in this chapter has stronger filtering ability and higher precision than the other two methods.

In order to analyze the effect of fan rotor blade leading edge erosion on the aerodynamic characteristics of the compression system for different internal and external culverting conditions, multilevel cascade calculations were used to investigate the changes in the aerodynamic characteristics of the compression system of a certain type of engine with large culverting ratio. The results show that the total pressure ratio, isentropic efficiency and channel flow rate of the engine compression system after leading edge erosion relatively decrease by 0.18%, 0.879% and 0.972%, respectively, under the peak efficiency point working condition. For fan rotor blades, leading edge erosion leads to a decrease in the slope of the surface static pressure coefficient curve in the axial 0~24% interval of the surface static pressure coefficient curve at the height of 95% of the blades under near-surge point conditions in the outer culvert channel, with a decrease in static pressure, and an increase in the slope in the axial 24%~40% interval, which reduces the loading of the blade, reduces the pressurizing capacity, and moves the surge wave forward.In addition, leading edge erosion increases the angle of attack of the overall fan blade at the near-surge point of the outer culvert by about 2°, resulting in changes in blade aerodynamic efficiency and stall margin, which deviates from the original blade design.

The dynamic modulus of asphalt mixture is an important parameter in the design of asphalt pavement. Extracting material characteristics, dynamic modulus, and phase angle information from a large amount of asphalt concrete datasets using integrated methods is of great significance for optimizing the performance of asphalt pavement. The extreme gradient boost (XGBoost) model aggregated a series of decision tree models through weighted summation to construct a powerful prediction model, while optimizing the loss function to minimize prediction errors. In order to further improve the accuracy of dynamic modulus and phase angle prediction, heuristic algorithms were used to optimize the model. Initially, the basic model was initialized based on samples and the gradient of the loss function of the training data was calculated. Subsequently, XGBoost utilized gradient details to construct a decision tree model, optimized leaf node weights, and updated the model’s predictions through weighted summation. During this process, heuristic algorithms are used to optimize the optimal parameters of the entire XGBoost model. The experimental results show that the improved XGBoost model outperforms the original model in all performance evaluation indicators, improving the accuracy of predicting the dynamic modulus and phase angle of asphalt mixtures.

The accuracy and confidence level of the unmanned aerial vehicle(UAV) landing range are of great significance for objectively assessing the UAV’s ground risk. The uncertain wind field and complex electromagnetic environment are the main causes of uncertainty regarding UAV failure and landing range. Given the particle assumption in the case of complete failure of the UAV, firstly, a dynamic model of the UAV trajectory descent with the initial position and velocity as the boundary value and the wind speed vector and the initial position as random variables was constructed, and the failure and landing range of the UAV were determined by Monte Carlo simulation. Secondly, a geometric method for determining the envelope of the UAV ground risk buffer was proposed, and the quantitative determination of the ground risk buffer of the entire UAV track was realized. Finally, the method proposed was verified by taking an aerial inspection route as example and compared with the buffer protection area of the UAV in different wind fields and under various operating conditions, the effect of uncertain wind field and its operating conditions on the ground crash range of the UAV was studied, and the ground risk buffer zone under different operating conditions was established. The results show that falling at higher speed and higher altitudes under stronger winds yields a wider impacting area and a larger ground risk buffer.

To achieve rapid and reliable prediction of asphalt mixture performance, a method for predicting asphalt mixture performance by optimizing the back propagation (BP) neural network with a genetic algorithm (GA) from the perspective of material composition design was proposed. Initially, a grey relational analysis method was employed to reduce the dimensionality of multidimensional input variables, identifying the core influencing factors of asphalt mixture performance. Subsequently, integrating the GA, a GA-BP neural network prediction model was constructed with the core influencing factors as the input layer and asphalt mixture performance as the output layer. The model underwent training, validation analysis, and prediction generalization application. A comparison with the training effectiveness and prediction accuracy of the BP neural network was conducted to verify the accuracy of the GA-BP neural network model. The research results indicate that the grey relational degrees of eight performance characteristics, including air void, asphalt-aggregate ratio, nominal maximum aggregate size, 4.75 mm passing rate, asphalt type, softening point, penetration, and ductility, are all greater than 0.6, signifying their significant impact on asphalt mixture performance. Compared to the BP neural network model, the GA-BP neural network model reduces the root mean square error (RMSE) by 16% to 31%, decreases the mean absolute error (MAE) by 15% to 24%, and improves the R² value by 0.01 to 0.27, indicating that it has better learning and fitting capabilities. The prediction accuracy for dynamic modulus, dynamic stability, residual stability, splitting tensile strength ratio, and ultimate bending strain of the asphalt mixture is respectively enhanced by 35.26%, 47.78%, 23.13%, 31.92%, and 35.75%, revealing the superior generalization application capability of the GA-BP neural network model. The research findings provide essential references for the rapid prediction of asphalt mixture performance and guidance in the design of asphalt mixture material composition.

The creep of wood under the action of long-term load will increase the deformation of wooden beams or wooden columns, and bring hidden safety risks to the building structure. In order to improve the creep performance of wood components, an aluminum wood composite columns (AWC) was designed to make AWC with 0,2%, 3%, 3% and 4% respectively. The creep test of 0.25 stress ratio of the lower column lasted 30 days. The creep strain-time curve and creep coefficient-time curve were obtained and the AWC creep pattern was analyzed. Burgers model was used to fit the creep strain-time curve to explore the influence of aluminum alloy content on AWC creep, and analyze the causes of AWC int creep inhibition from AWC material characteristics and load transfer. The results show that, compared with AWC 1 with aluminum alloy content of 0, the creep deformation of AWC 2 ~ AWC 4 decreases to different degrees, that is, aluminum alloy can effectively enhance the creep deformation ability of wood; The four AWC creep strain-time curve correlation coefficient based on burgers model are greater than 0.95, and the creep prediction model constructed by this model can predict the AWC creep.

To enhance the synchronization performance of multiple clocks in robotic communication systems, a study was conducted on a distributed clock synchronization compensation scheme. Quantitative compensation and dynamic compensation algorithms for clock errors were designed. The synchronization compensation scheme, based on the IgH EtherCAT communication protocol stack, was implemented, and trajectory tracking and communication performance tests were conducted on a six-axis robotic arm. The results show that the application of this synchronization scheme improves the circular trajectory precision of the robotic arm’s end effector compared to non-IgH synchronization schemes. The clock synchronization error is reduced to 54 ns, with the error tolerance within ±200 ns. It is evident that the compensation scheme enhances clock synchronization performance and meets the application requirements of the robotic arm.

The classical earth pressure theory only calculates the ultimate earth pressure in translational mode of the retaining wall. In order to get closer to the real engineering situation, it is necessary to develop the theory of non-ultimate earth pressure of the cohesive fill behind the wall under arbitrary load action and arbitrary displacement mode. The concept of equivalent uniform load was adopted to convert any type of load acting on the earth filling surface into uniform load. Based on the classical earth pressure theory, the rigid retaining wall was regarded as a combination of an ideal rigid plastic body and a series of springs, and the soil stiffness coefficient of the clay was obtained, which improved the Coulomb earth pressure theory and derives the general calculation method of the non-ultimate passive earth pressure of the soil in any displacement mode that can consider the arbitrary load.The research results indicate that the distribution and magnitude of passive soil pressure, the position of the resultant force point are closely related to the displacement mode of the retaining wall, the magnitude of cohesive force, and the load acting on the fill surface. When the equivalent uniform load and cohesive force are both 0, this method degenerates into Coulomb soil pressure theory.The model validation results show that the theoretical calculated values are in good agreement with the experimental measured values, proving that this calculation method has certain theoretical significance for calculating the passive soil pressure of cohesive soil under non limit states, and also has practical value for engineering practice.