In the intelligent optimization process of aerospace thin-walled parts, there are issues such as solidification of core knowledge base, high system coupling degree, and real-time evaluation and optimization feedback required for the knowledge base. These problems make it difficult to expand the functions of the digital twin system and meet the growing processing needs, ultimately hindering the application of digital twin technology. To address these issues, a digital twin system for controlling processing errors in thin-walled parts was built using a microservices architecture. In addition, a method for building a digital twin system at the processing unit level with the best coupling degree was proposed, mainly targeting the dynamic characteristics analysis knowledge base of thin-walled parts. Furthermore, to meet the requirements for backward compatibility of the processing unit level digital twin system, a comprehensive solution including the construction, operation, evaluation, optimization, and visualization of a knowledge base for the dynamic characteristics of the processing unit was proposed, providing guidance for the digital transformation and upgrading of CNC machine tools and the optimization of processing technology based on digital twin technology.
A TC4 impeller blade is a typical weak, rigid, thin–walled part. The contact area between a cutting tool and a workpiece has strong time–varying characteristics. This leads to a strong non–linear variation in cutting load. So, in this kind of part, the processing error is difficult to control. To solve this problem, a method of processing error prediction and intelligent controlling which considers the effect of tool wear time variation is proposed by combining digital–twinning technology. Firstly, an iterative model for digital–twin process optimization is constructed. Secondly, an iterative prediction model of the machining position following the milling force and considering the effect of tool wear is proposed. Based on these models, the machining error of the TC4 impeller blade under dynamic load is predicted. Dynamic machining error prediction and intelligent control are realized by combining the digital–twin model and the multi–objective process algorithm. Finally, the machining error optimization effect of the proposed digital–twin model is verified via a comparison experiment of impeller blade milling. In terms of the precision of milling force mapping, the average error after optimization is less than 8%. The maximum error is no more than 14%. In terms of the optimization effect, the average error of the optimized workpiece contour is reduced by about 20%. The peak contour error is reduced by approximately 35%.
Unmanned aerial vehicle (UAV)-based object detection shows promising potential in intelligent transportation and disaster response. However, detecting small targets remains challenging due to inherent limitations (long-distance and low-resolution imaging) and environmental interference (complex backgrounds and occlusions). To address these issues, this paper proposes an enhanced small target detection model, DDF-YOLO, which achieves higher detection performance. First, a dynamic feature extraction module (C2f-DCNv4) employs deformable convolutions to effectively capture features from irregularly shaped objects. In addition, a dynamic upsampling module (DySample) optimizes multi-scale feature fusion by combining shallow spatial details with deep semantic features, preserving critical low-level information while enhancing generalization across scales. Finally, to balance rapid convergence with precise localization, an adaptive Focaler-ECIoU loss function dynamically adjusts training weights based on sample quality during bounding box regression. Extensive experiments on VisDrone2019 and UAVDT benchmarks demonstrate DDF-YOLO’s superiority. Compared to YOLOv8n, our model achieves gains of 8.6% and 4.8% in mAP50, along with improvements of 5.0% and 3.3% in mAP50-95, respectively. Furthermore, it exhibits superior efficiency, requiring only 7.3 GFLOPs and attaining an inference speed of 179 FPS. These results validate the model’s robustness for UAV-based detection, particularly in small-object scenarios.
While CH3NH3PbBr3 (MAPbBr3) single crystals exhibit exceptional potential for biased X/γ-ray detection, achieving a stable metal–semiconductor contact remains a critical challenge. Herein, we address this issue by incorporating AgBr as a dopant into pristine MAPbBr3 single crystals prior to Ag electrode deposition. Our results reveal a concentration-dependent interfacial mechanism: while low-level doping introduces excess Br− ions and Pb vacancies that accelerate electrochemical corrosion, a high concentration of Ag+ within the lattice effectively suppresses the deleterious reaction between the Ag electrode and the perovskite. This work establishes a defect-engineering strategy to stabilize metal electrodes, offering a viable pathway for the development of reliable, high-performance MAPbBr3-based radiation detectors.