Binary gas mixture adsorption equilibrium data for the ethylene-carbon dioxide system were obtained for cation exchanged forms of ZSM5 (Li^+, Na^+, K^+, Rb^+, Mg^(+2), Ca^(+2), Sr^(+2), and Ba^(+2)) for the gas phase ...Binary gas mixture adsorption equilibrium data for the ethylene-carbon dioxide system were obtained for cation exchanged forms of ZSM5 (Li^+, Na^+, K^+, Rb^+, Mg^(+2), Ca^(+2), Sr^(+2), and Ba^(+2)) for the gas phase CO_2 mole fracion of 0.766 at 308K and 101. 3kPa. The experimental adsorption phase diagrams were obtained for CO_2-C_2H_4 on NaZSM5 and MgZSM5. Single component adsorption isotherms for CO_2 and C_2H_4 were also obtained for these two zeolites. The single component data were used to obtain parameters derived in the vacancy solution model (VSM) and the statistical thermodynamic model(STM). These parameters were, in turn, used to predict binary mixture isotherms for these two zeolites. The agreement between experimental data and predicted value is generally good.展开更多
为解决在光线昏暗、声音与视觉噪声干扰等复杂条件下,单模态鱼类行为识别准确率和召回率低的问题,提出了基于声音和视觉特征多级融合的鱼类行为识别模型U-FusionNet-ResNet50+SENet,该方法采用ResNet50模型提取视觉模态特征,通过MFCC+Re...为解决在光线昏暗、声音与视觉噪声干扰等复杂条件下,单模态鱼类行为识别准确率和召回率低的问题,提出了基于声音和视觉特征多级融合的鱼类行为识别模型U-FusionNet-ResNet50+SENet,该方法采用ResNet50模型提取视觉模态特征,通过MFCC+RestNet50模型提取声音模态特征,并在此基础上设计一种U型融合架构,使不同维度的鱼类视觉和声音特征充分交互,在特征提取的各阶段实现特征融合,最后引入SENet构成关注通道信息特征融合网络,并通过对比试验,采用多模态鱼类行为的合成加噪试验数据验证算法的有效性。结果表明:U-FusionNet-ResNet50+SENet对鱼类行为识别准确率达到93.71%,F1值达到93.43%,召回率达到92.56%,与效果较好的已有模型Intermediate-feature-level deep model相比,召回率、F1值和准确率分别提升了2.35%、3.45%和3.48%。研究表明,所提出的U-FusionNet-ResNet50+SENet识别方法,可有效解决单模态鱼类行为识别准确率低的问题,提升了鱼类行为识别的整体效果,可以有效识别复杂条件下鱼类的游泳、摄食等行为,为真实生产条件下的鱼类行为识别研究提供了新思路和新方法。展开更多
文摘Binary gas mixture adsorption equilibrium data for the ethylene-carbon dioxide system were obtained for cation exchanged forms of ZSM5 (Li^+, Na^+, K^+, Rb^+, Mg^(+2), Ca^(+2), Sr^(+2), and Ba^(+2)) for the gas phase CO_2 mole fracion of 0.766 at 308K and 101. 3kPa. The experimental adsorption phase diagrams were obtained for CO_2-C_2H_4 on NaZSM5 and MgZSM5. Single component adsorption isotherms for CO_2 and C_2H_4 were also obtained for these two zeolites. The single component data were used to obtain parameters derived in the vacancy solution model (VSM) and the statistical thermodynamic model(STM). These parameters were, in turn, used to predict binary mixture isotherms for these two zeolites. The agreement between experimental data and predicted value is generally good.
文摘为解决在光线昏暗、声音与视觉噪声干扰等复杂条件下,单模态鱼类行为识别准确率和召回率低的问题,提出了基于声音和视觉特征多级融合的鱼类行为识别模型U-FusionNet-ResNet50+SENet,该方法采用ResNet50模型提取视觉模态特征,通过MFCC+RestNet50模型提取声音模态特征,并在此基础上设计一种U型融合架构,使不同维度的鱼类视觉和声音特征充分交互,在特征提取的各阶段实现特征融合,最后引入SENet构成关注通道信息特征融合网络,并通过对比试验,采用多模态鱼类行为的合成加噪试验数据验证算法的有效性。结果表明:U-FusionNet-ResNet50+SENet对鱼类行为识别准确率达到93.71%,F1值达到93.43%,召回率达到92.56%,与效果较好的已有模型Intermediate-feature-level deep model相比,召回率、F1值和准确率分别提升了2.35%、3.45%和3.48%。研究表明,所提出的U-FusionNet-ResNet50+SENet识别方法,可有效解决单模态鱼类行为识别准确率低的问题,提升了鱼类行为识别的整体效果,可以有效识别复杂条件下鱼类的游泳、摄食等行为,为真实生产条件下的鱼类行为识别研究提供了新思路和新方法。
