Two different types of dual phase steels DP780 were manufactured and investigated. Steel A contained higher carbon, the other one B had lower carbon with Nb added. The mechanical properties and ...microstructures were analyzed and compared. The hole expansion test were conducted to evaluate stretch-flangeability. An Ultra Micro-Indentation System (UMIS) were employed to carry out the nano-hardness test to check the nano-hardness of ferrite and martensite. The result showed that for DP steel A, higher over-aging temperature during manufacturing would result in lower ultimate tensile strength with martensite partly resolved, which would lead to higher stretch-flangeability. In comparison with A, steel B with lower carbon and Nb added exhibited higher stretch-flangeability, which was related to the refined grain size of ferrite. It could be concluded that for dual phase steel stretch-flangeability was mainly affected by the strength difference between ferrite and martensite. Generally the bigger the strength difference is, the lower the stretch-flangeability is. Either decreasing martensite strength by increasing over-aging temperature or increasing ferrite strength by adding Nb could improve the stretch-flangeability. The same parts formed from the two different dual phase steel were compared. There was crack in some edge area during forming for steel A, while the part formed from Steel B showed good quality without crack, which verified the stretch-flangeability improvement was very effective. However, increasing over-aging temperature during manufacturing is not easy to be implemented and it could result in lower ultimate tensile strength, which may not meet the strength requirement. Adding Nb is considered to be a more feasible and effective way to improve ferrite strength and stretch-flangeability for dual phase steel DP780.
An effective battery thermal management system (BTMS) of power battery module for electric vehicles (EVs) plays a decisive role in battery life, cost, and safety in use. A BTMS method for a ...lithium-ion battery module, which is consisting of 12 prismatic cells, is proposed based on liquid cooling in the present work. Three circuitous channels embedded with a cooling plate are studied. In addition, the reliability and scientificity of the numerical model are verified by cooling performance tests. The simulation and test for this BTMS under a 3C discharging and charging rate were performed. Meanwhile, the performance evaluations of the battery module by considering different cooling conditions and battery arrangement are investigated numerically. With the help of the numerical model and test, the effects of flow area, mass flow rate, and cell gap on the battery temperature are examined and discussed. Simulation results illustrate that the larger flow area is experienced with a lower maximum temperature. BTMS with five channels design exhibits better heat dissipation behavior, and the maximum temperature and temperature difference under 80 g/s mass flow rate is about 45°C. Then, increases of maximum temperature and temperature difference are shown under lower mass flow rate, and the peak temperature at the lowest flow rate of 2 g/s is equal to 79.89°C, while it increases to 80 g/s, the peak temperature drops down to 47.43°C. Moreover, the lower the cell gap is, the worse the cooling performance is. Consequently, this study of this paper will provide a theoretical and an experimental basis for improving the battery module thermal behavior and BTMS strategy.
With the rapid development of technologies such as smart home, smart medical and autonomous driving, lightweight networks play an important role in promoting the application of deep learning ...technology on mobile and embedded terminals. The Transformer-based model changes the architecture of traditional neural networks, and performs outstanding in many fields such as natural language processing and computer vision. However, the huge computing cost and the increasing network scale have increased the demand for storage, running memory and computing, which hinders their widespread deployment on various hardware devices, such as mobile phones, robots and Internet of Things (IoT) devices. Therefore, compression method for Transformer-based models need to be explored so that they can be applied more widely on mobile devices. In this paper, we focus on the lightweight and optimization acceleration methods on Vision Transformer in recent years, and summarize them as quantization, knowledge distillation, pruning and adjusting network structure, the innovations, merits and demerits of these approaches have been compared and reviewed. Through this survey, it is hoped to provide useful help for the current compression method for Vision Transformer, and also expected to point out a direction for future research in this field.
In the process of producing and manufacturing leather, a lot of unavoidable damage will be caused by complex technology, so that there are various defects on the surface. In order to control the ...quality of the product and guide the subsequent processing, the location and removal of the defective part have become an important part of the production process. The traditional manual inspection method is time consuming, low efficiency and easy to make mistakes. To address the above issues, a novel Defect Detection method, LD2-YOLO, is proposed for automotive composite Leather in this paper. Firstly, the LD2-YOLO is built upon YOLOv7 architecture, and a new feature extraction module HorBlock is introduced into the model to learn global features by replacing several convolution layers. In addition, global filter is adopted for the HorBlock to extract frequency domain features. Therefore, both the spatial and frequency domain features can be generated by the model. Secondly, the training process is optimized by the GELU activation function to provide faster and better convergence. Finally, the validity of LD2-YOLO is verified on our self-collected Leather Defect Dataset (LDD). The experimental results show that the average detection accuracy of LD2-YOLO can reach 99.1% with a small increase of model parameters and GFLOPs, which is better than the baseline.
Online Florist System Based on JavaWeb Chen, Meng; Liu, Tao; Yu, Wuxin ...
2021 2nd International Conference on Artificial Intelligence and Computer Engineering (ICAICE),
2021-Nov.
Conference Proceeding
The sudden COVID-19 pandemic made us feel the limitations of offline sales. In order to let consumers fully realize the convenience of the Internet era and highlight the advantages of online sales, a ...personalized online florist system (OFS) is developed. The website takes JSP (JavaServer Pages) + SSM (Spring + SpringMVC + MyBatis) framework as the development technology and based on B/S (Browser/Server) structure, an online florist system based on JavaWeb is designed and implemented.
The residual stress in cold rolling could decrease the accuracy of ring parts. Because many factors could cause residual stress, it’s difficult to calculate it theoretically. At present, a ...measurement method was generally used to study the residual stress. For example, X ray could be used to measure residual stress in rings. Through the measurement of residual stress, it could not only obtain the residual stress level after ring rolling, but also validate the finite element model to predict the residual stress distribution. However, because the ring surface was circle, it needs to cut a plane for X ray measure, and it increases measurement inaccuracy. This method only could be suited for qualitative analysis of residual stress in cold ring rolling, and it is not suitable for quantitative analysis of residual stress distribution.