Among the numerous novel eco-friendly insulating gases, C 4 F 7 N has attracted much attention due to its excellent electrical performance. However, except for the electrical perfomance, the compatibility between the gas medium and the sealing materials is equally important for gas-insulated equipment. At present, studies about the compatibility between C 4 F 7 N and EPDM, a widely used sealing material in power systems, are available in some previous works, but few focused on the compatibility comparison between C 4 F 7 N gas mixtures and EPDM with different third monomers. In this paper, we carried out the thermal aging test on ENB-EPDM, DCPD-EPDM, and C 4 F 7 N gas mixture to perfect the compatibility mechanism between EPDM and C 4 F 7 N. It was found that both of the EPDM reacted with the gas mixture and led to the property changes in the solid samples and the decomposition of C 4 F 7 N. On the other hand, by coating silicone grease, the contact between gas and rubber was effectively blocked and the concentration of the decomposition product was significantly reduced. The performance comparison indicates that ENB-EPDM is more suitable for sealing the C 4 F 7 N gas mixture, which is due to the superior thermal stability of ENB.
In multi-task learning, difficulty levels of different tasks are varying. There are many works to handle this situation and we classify them into five categories, including the direct sum approach, the weighted sum approach, the maximum approach, the curriculum learning approach, and the multi-objective optimization approach. Those approaches have their own limitations, for example, using manually designed rules to update task weights, non-smooth objective function, and failing to incorporate other functions than training losses. In this paper, to alleviate those limitations, we propose a Balanced Multi-Task Learning (BMTL) framework. Different from existing studies which rely on task weighting, the BMTL framework proposes to transform the training loss of each task to balance difficulty levels among tasks based on an intuitive idea that tasks with larger training losses will receive more attention during the optimization procedure. We analyze the transformation function and derive necessary conditions. The proposed BMTL framework is very simple and it can be combined with most multi-task learning models. Empirical studies show the state-of-the-art performance of the proposed BMTL framework.
We use the proper orthogonal decomposition (POD) method to decompose acoustic imaging from wavelet-based beamforming results into their corresponding modes, and demonstrate the method to analyze the noise characteristics of small-dimensional components, where research literature is still scarce. In particular, we consider the noise from a central processing unit (CPU) cooling fan, though the developed method could find realistic applications in a wider scheme of electronic industry. In addition to developing the new analysis method, we also endeavor to prepare a realistic experimental setup with a CPU cooling fan in an anechoic room. The results show that the combination of POD and wavelet-based beamforming methods is suitable for CPU cooling fan acoustic imaging, which can enhance the analysis ability of acoustic imaging test. The results offered in this work are helpful in cooling fan design and evaluation, which would open doors to a wide range of industrial applications. INDEX TERMS Acoustic imaging, proper orthogonal decomposition, fan noise, wavelet.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.