In today’s rapid widespread of digital technologies into all live aspects to enhance efficiency and productivity on the one hand and on the other hand ensure customer engagement, personal data counterfeiting has beco...In today’s rapid widespread of digital technologies into all live aspects to enhance efficiency and productivity on the one hand and on the other hand ensure customer engagement, personal data counterfeiting has become a major concern for businesses and end-users. One solution to ensure data security is encryption, where keys are central. There is therefore a need to find robusts key generation implementation that is effective, inexpensive and non-invasive for protecting and preventing data counterfeiting. In this paper, we use the theory of electromagnetic wave propagation to generate encryption keys.展开更多
The equivalent filter modeling of a rectangular dielectric post in a rectangular waveguide is obtained through the variational expression of input impedance. The reflection coefficient expressed in components of netwo...The equivalent filter modeling of a rectangular dielectric post in a rectangular waveguide is obtained through the variational expression of input impedance. The reflection coefficient expressed in components of network is in good agreement with the results given by K. Siakavara, et al. (1991), The method can be applied to design filter.展开更多
界面面积浓度(Interfacial Area Concentration,IAC)是两相流动的封闭两流体模型中界面传递项的关键参数,用于表征气液界面传输能力的强弱。对界面面积浓度的建模预测通常有经验关联式和界面面积输运方程等方法,但这些方法都有较大的数...界面面积浓度(Interfacial Area Concentration,IAC)是两相流动的封闭两流体模型中界面传递项的关键参数,用于表征气液界面传输能力的强弱。对界面面积浓度的建模预测通常有经验关联式和界面面积输运方程等方法,但这些方法都有较大的数据依赖性。而对神经网络模型增添可解释性可以为模型修正提供方向,严谨提升预测精度。为更好地预测两相流动的IAC,基于神经网络建立了IAC的预测模型,结合不同气泡行为、物理关系及统计分布并利用事后可解释性方法,对比分析了不同输入特征组合的神经网络模型预测能力;并通过神经网络每层的结构参数大小,分析输出比重挑选合适的数据预处理方法。通过事后可解释性分析得到空泡份额是IAC预测的重要因素,而对训练数据进行对数变换预处理能够显著提高模型对真实数据的预测能力。展开更多
文摘In today’s rapid widespread of digital technologies into all live aspects to enhance efficiency and productivity on the one hand and on the other hand ensure customer engagement, personal data counterfeiting has become a major concern for businesses and end-users. One solution to ensure data security is encryption, where keys are central. There is therefore a need to find robusts key generation implementation that is effective, inexpensive and non-invasive for protecting and preventing data counterfeiting. In this paper, we use the theory of electromagnetic wave propagation to generate encryption keys.
文摘The equivalent filter modeling of a rectangular dielectric post in a rectangular waveguide is obtained through the variational expression of input impedance. The reflection coefficient expressed in components of network is in good agreement with the results given by K. Siakavara, et al. (1991), The method can be applied to design filter.
文摘界面面积浓度(Interfacial Area Concentration,IAC)是两相流动的封闭两流体模型中界面传递项的关键参数,用于表征气液界面传输能力的强弱。对界面面积浓度的建模预测通常有经验关联式和界面面积输运方程等方法,但这些方法都有较大的数据依赖性。而对神经网络模型增添可解释性可以为模型修正提供方向,严谨提升预测精度。为更好地预测两相流动的IAC,基于神经网络建立了IAC的预测模型,结合不同气泡行为、物理关系及统计分布并利用事后可解释性方法,对比分析了不同输入特征组合的神经网络模型预测能力;并通过神经网络每层的结构参数大小,分析输出比重挑选合适的数据预处理方法。通过事后可解释性分析得到空泡份额是IAC预测的重要因素,而对训练数据进行对数变换预处理能够显著提高模型对真实数据的预测能力。