By predicting the wear trend of aeroengine, the wear state of aeroengine can be monitored effectively. Among the effective observation data reflecting the engine wear state, the oil analysis data can indirectly reflect the overall wear trend of aeroengine. Therefore, by establishing a trend prediction model based on oil sample analysis data, so as to realize the wear trend prediction of engine. However, the current models used in aeroengine trend prediction are mainly single prediction models, and the combined prediction models are only general linear combinations, with poor prediction effect. Therefore, a nonlinear variable weight combination prediction model based on support vector machine is proposed, and realizes the parameter optimization through particle swarm optimization algorithm. The oil sample analysis data is obtained through the bearing fatigue test of the whole life oil system, and the oil samples are collected at fixed intervals for performance analysis. Through the combination prediction analysis of the spectral analysis data, by comparing the prediction results of the combination prediction and the prediction results of the single prediction model, the prediction accuracy exceeds the prediction accuracy of the single prediction model, which fully verifies the superiority and effectiveness of the combination prediction model proposed in this paper.

During the operation of an air cushion belt conveyor, factors such as fan air volume, number of air holes, and film thickness have a significant impact on the flow field characteristics and bearing capacity of the air cushion. By establishing a simulation model for the air cushion flow field, analyze the changes in air cushion pressure under different working conditions. Based on experimental comparison and analysis, the variation law of air cushion pressure under different air volume and number of air holes, as well as the variation of air film thickness and air film pressure under different air volume were analyzed. The results show that as the air volume increases, the film pressure gradually increases, and the film thickness changes from 5% to 30%. As the number of pores increases, the pressure gradient of the air cushion changes faster and exhibits a parabolic distribution, and the degree of change in air film thickness decreases. The experimental and simulation results show that the changes in the air cushion flow field are consistent, and the fan air volume is 15-20 m³/m, the optimal value K₁ for the stability of the air cushion flow field is 1.31, and the optimal working condition is 5 exhaust holes, which meets the requirements of actual operating conditions.

In the railway container yard, there are few mature intelligent anti-lifting solutions available for train flatbed loading and unloading operations due to the poor detection accuracy or speed of traditional detection methods. This paper proposes a fast anti-lifting detection method for trains based on an improved back propagation (BP) neural network. By acquiring weight data from the four locks of the hoist, a flatbed lifting detection model is established using a BP neural network. During weight adjustment, a momentum factor and an adaptive learning rate are incorporated to optimize the model's performance. Through practical tests, this method demonstrates that this model achieves a high detection rate and fast detection speed, making it suitable for providing intelligent safety protection for automated rail mounted gantry in the railway container yard.

To thoroughly explore the extensive applications and prospects of knowledge graphs in the domain of intelligent manufacturing, aiming to support the sustained development of the manufacturing industry, the domain of intelligent manufacturing has been categorized into four dimensions: vertical industry applications, manufacturing process applications, domain graph construction technology, and intelligent services. Through this study, the significance of knowledge graphs in driving the evolution of intelligent manufacturing is reviewed. Furthermore, a framework for an intelligent manufacturing knowledge graph, rooted in manufacturing domain knowledge data, is proposed. This framework encompasses three key modules: manufacturing domain data, graph construction, and intelligent services, providing theoretical support for the continuous upgrading of intelligent manufacturing. Research findings emphasize the crucial role of knowledge graphs in advancing the intelligence of the manufacturing industry and the broad potential of knowledge graphs in the field of intelligent manufacturing. Additionally, an outlook on future research directions for knowledge graphs in the domain of intelligent manufacturing is suggested, with a focus on exploring the integration of domain graphs with next-generation artificial intelligence technologies to propel the continuous innovation and intelligent evolution of manufacturing.

A theoretical model of stress distribution and contact width is established by Hertz contact theory. The reaction of roller under specific load is analyzed. The results shows that linear profile have significant edge contact stress concentration, which increases with the increase of load. The roller with logarithmic profile can solve the stress concentration, but it can't completely eliminate it. Under the same load, with the increase of the bias factor, the maximum contact stress of the roller decreases sharply at first and then increases slowly. The position where the maximum contact stress occurs shifts from both ends of the roller to the middle of the roller. The profile design of roller with logarithmic type can take a value between 1.0 and 2.5 for the bias factor based on working conditions, roller size and load conditions. The value of bias factor is adaptive to tilt error and the impact of tilt error on contact stress can be reduced by setting a reasonable value.

In order to investigate the effect of tower slewing motion on tip displacement and to provide a basis for collecting tip displacement data to improve the accuracy of tower steel structure damage diagnosis, this paper adopts a tower physical structure experiment bench to select three different strokes of high, medium and low slewing speeds, and to collect tip displacement during slewing motion and static tip displacement. A characterization study of the tip displacement for slewing stroke and speed change was carried out; the structural response characteristics of the slewing motion were established and characterized by noise characterization coefficients. The experimental results show that the collected tip displacements are close to static displacements when the slewing stroke is larger, the slewing speed is slower, and the noise characterization coefficient is smaller than a set threshold. The purpose of this study is to avoid the influence of slewing motion on the real-time monitoring of structural damage, and to give a scientific basis for the setting of the data acquisition conditions for the required tower steel structure damage diagnosis.

In shield construction, the tunneling attitude of the shield tunnel is an important parameter affecting the quality of the tunneling, which directly affects the track and the quality of the tunnel. Taking an earth pressure balance tunnel boring machine (TBM) as an example, the relationship between the weight of the main machine and the earth pressure and the bearing capacity of the foundation is determined from the theoretical level at the design of the shield machine, which is based on the analysis of the floating and sinking. The influence of the uneven mass distribution on the center of gravity is further analyzed. Finally, through the comprehensive study of the relationship between the force of the host and the thrust of the cylinder under the soil pressure model, the necessary conditions for the control of the driving attitude of the shield machine are determined. The results of this paper have important practical siginificance for improving the design and manufacturing of the shield machine and the construction level of the shield method in China.

Lightweight design is an effective way to improve the economic and environmental protection of static pile driver, as the key component of the static pile driver, the combined fixture is important for its lightweight design. To realize the combined fixture structure safety and lightweight design goal, a three-dimensional model is established by using solid works, and ANSYS Workbench is used to analyze and evaluate the equivalent stress of the whole structure and main components under the condition of pile pressing. Based on the results of analysis and evaluation, the topology optimization method of superposition replacement is adopted to carry out the structural optimization design, while meeting the needs of safety and engineering practice, its total quality is reduced by about 30%, and the economic and environmental benefits are obvious.

This paper proposes a robotic grasping technique based on object recognition and fully convolutional grasp quality convolutional neural network (FC-GQCNN). To address the limitations of traditional GQCNN, such as low computational efficiency and redundant feature calculations, an improved FC-GQCNN is developed. By replacing the fully connected layers in GQCNN with 1×1 convolutional layers, the proposed network can handle input images of arbitrary sizes. Furthermore, the integration of FC-GQCNN with the YOLOv8 object detection algorithm forms a YOLOv8-FCGQCNN cascade structure, effectively solving the challenges of object recognition and localization in complex environments. Experimental results demonstrate that this method achieves an 86% grasp success rate across 10 different objects, with an average detection time of 0.09 s per frame, which is 22 times faster than traditional GQCNN, significantly improving system efficiency. This method can accurately detect the grasping position of the object of interest and has higher reliability than the baseline method.

In order to solve the problems of insufficient monitoring strength, poor interaction and low digitization degree during the operation of mobile crane, the construction method of digital twin system for real-time monitoring of crane operation status was proposed. A five-dimensional digital twin model is introduced to establish a digital twin framework for virtual and real control of cranes. Build a physical space from the service scene and physical entity; Establish virtual space from virtual model and visual scene and action control. Based on MySQL database, the twin database is constructed by using inherent, collected and virtual information. Virtual and real interaction, dynamic monitoring and visualization are realized in combination with communication protocols. Taking YDC20/30 light and small mobile crane as an example, the feasibility of this method is verified, and a new scheme is provided for comprehensively controlling the service process of the crane.