频率转移和反向设计在多物理合下通过欧勒潜伏动态和数据分析规范化的元表面
Enze Zhu1, Zheng Zong1, Erji Li1
1Innovation Institute of Electromagnetic Information and Electronics Integration, Zhejiang Key Laboratory of Intelligent Electromagnetic Control and Advanced Electronic Integration, College of Information Science & Electronic Engineering, Zhejiang University, Hangzhou, China.
Nature communications
|March 6, 2025
概括
这项研究引入了一种新的多物理深度学习框架 (MDLF),用于在机器学习中准确的频率转移. 该框架允许对超表面进行范围外预测,而不需要广泛的频率特定训练数据.
科学领域:
- 计算物理学的计算物理.
- 机器学习是机器学习.
- 材料科学 是一种材料科学.
背景情况:
- 频率转移对于超出训练数据范围的机器学习预测至关重要.
- 传统的深度神经网络 (DNN) 需要特定的频率数据,这对于多物理问题来说往往是无法访问的.
- 测量或计算的局限性阻碍了在某些频率上获取数据.
研究的目的:
- 开发一个通用的深度学习框架,用于在地表分析中准确的频率转移.
- 为了在没有全面的频率特定训练数据的情况下实现范围外预测.
- 将混合先前信息的反转方法纳入 metasurface反向设计中.
主要方法:
- 提出了一个多物理深度学习框架 (MDLF),集成了一个多忠实度的DeepONet,一个欧勒潜伏动态网络和一个数据分析反转网络.
- 该框架动态地利用欧勒隐性空间和单一物理信息来概括到看不见的频段.
- 引入了一个反向方法,以结合混合先验信息来实现反向设计.
主要成果:
- MDLF成功地将参数和自由形态元表面的频段泛化为未见的频段.
- 该框架表现出有效的预测能力,而不需要先前了解多物理反应.
- 在电磁热合下进行的数值和实验验证证证了MDLF的有效性.
结论:
- 拟议的MDLF为机器学习中的频率转移挑战提供了强大的解决方案,特别是对于多物理问题.
- 这一框架显著扩大了深度学习模型对超表面设计的预测能力.
- 该研究通过严格的测试验证了MDLF在现实应用中的性能.
相关概念视频
01:23Transmission Line Design Considerations
113
Aluminum has become the material of choice for overhead transmission lines, surpassing copper due to its abundance and cost-effectiveness. The most prevalent type is the aluminum conductor, steel-reinforced (ACSR), which combines aluminum strands around a steel core. Other variants include all-aluminum conductors (AAC), all-aluminum alloy conductors (AAAC), aluminum conductor alloy-reinforced (ACAR), and aluminum-clad steel conductors. Advanced designs, such as aluminum conductors with steel...
113
01:12Mesh Analysis for AC Circuits
332
In the domain of radio communication, the significance of impedance matching must be considered. It is crucial to ensure the efficient transmission of signals between radio transmitters and receivers. Achieving this balance involves using impedance-matching circuits, with one fundamental configuration comprising a resistor, capacitor, and inductor.
The process of harmonizing these impedances begins with a clear understanding of the input and output signals. Once these signals are known, the...
The process of harmonizing these impedances begins with a clear understanding of the input and output signals. Once these signals are known, the...
332
01:24Properties of Fourier Transform II
150
The Fourier Transform (FT) is an essential mathematical tool in signal processing, transforming a time-domain signal into its frequency-domain representation. This transformation elucidates the relationship between time and frequency domains through several properties, each revealing unique aspects of signal behavior.
The Frequency Shifting property of Fourier Transforms highlights that a shift in the frequency domain corresponds to a phase shift in the time domain. Mathematically, if x(t) has...
The Frequency Shifting property of Fourier Transforms highlights that a shift in the frequency domain corresponds to a phase shift in the time domain. Mathematically, if x(t) has...
150
01:28Magnetostatic Boundary Conditions
850
An electric field suffers a discontinuity at a surface charge. Similarly, a magnetic field is discontinuous at a surface current. The perpendicular component of a magnetic field is continuous across the interface of two magnetic mediums. In contrast, its parallel component, perpendicular to the current, is discontinuous by the amount equal to the product of the vacuum permeability and the surface current. Like the scalar potential in electrostatics, the vector potential is also continuous...
850
01:25Network Function of a Circuit
251
Frequency response analysis in electrical circuits provides vital insights into a circuit's behavior as the frequency of the input signal changes. The transfer function, a mathematical tool, is instrumental in understanding this behavior. It defines the relationship between phasor output and input and comes in four types: voltage gain, current gain, transfer impedance, and transfer admittance. The critical components of the transfer function are the poles and zeros.
251
01:30Propagation Speed of Electromagnetic Waves
3.3K
Electromagnetic waves are consistent with Ampere's law. Assuming there is no conduction current Ampere's law is given as:
3.3K