A Mesh Space Mapping Modeling Method with Mesh Deformation for Microwave Components.

Area of Science:

• Electromagnetics and Microwave Engineering
• Computational Modeling and Simulation

Background:

• Accurate modeling of microwave components is crucial for device design and performance prediction.
• Existing modeling techniques often face trade-offs between computational efficiency and accuracy.
• Space mapping (SM) offers a framework to bridge coarse and fine models, but improvements are needed.

Purpose of the Study:

• To propose a low-cost space mapping (SM) modeling method with mesh deformation for microwave components.
• To enhance the accuracy and computational efficiency of electromagnetic response prediction.
• To reduce the training data and computational cost associated with microwave component modeling.

Main Methods:

• Developed a coarse-mesh model with embedded automatic mesh deformation.
• Utilized space mapping (SM) to establish a relationship between coarse and fine mesh models.
• Applied the mesh SM modeling technique to a four-pole waveguide filter for validation.

Main Results:

• Achieved training and test errors below 1%, outperforming existing artificial neural network (ANN) and SM models.
• Demonstrated significant computational savings, reducing CPU time by approximately 70% compared to HFSS for predicting 100 data points.
• The proposed model accurately represents fine model features with fewer training data and lower computational cost.

Conclusions:

• The proposed mesh deformation-based SM modeling technique offers a highly accurate and efficient solution for microwave component analysis.
• This method effectively balances computational cost and predictive accuracy, making it suitable for practical applications.
• The technique shows promise for accelerating the design and optimization of microwave devices with reduced computational resources.