Spiral inductor modeling method based on random forest

Abstract

The invention discloses a spiral inductor modeling method based on a random forest, and the method comprises the following steps: 1, obtaining structural parameters and corresponding performance parameters of a spiral inductor through ANSYS HFSS electromagnetic simulation software, and constructing an original data set of the spiral inductor; 2, performing normalization processing on the originaldata set of the spiral inductor; 3, randomly dividing the normalized original data set into an original training sample set and a test sample set; 4, training the training sample set by using a randomforest algorithm, and obtaining a trained spiral inductance prediction model by using structural parameters of the spiral inductance as input and performance parameters as output; and 5, inputting the test sample set into the trained spiral inductance prediction model to obtain corresponding spiral inductance performance parameters, and evaluating the prediction precision of the model. The spiralinductor modeling method based on the random forest is established, the development and optimization of spiral inductors by designers are simplified, the research efficiency is improved, the cost issaved, and the integrated circuit design is accelerated.

Description

technical field [0001] The invention relates to a random forest-based spiral inductance modeling method, which belongs to the field of wireless communication. Background technique [0002] Spiral inductors are one of the most commonly used passive components in RF IC design, and the lack of fast and accurate computer-aided design tools like CAD is still one of the bottlenecks in RF circuit design. To meet the design requirements of RF on-chip spiral inductors, fast and accurate circuit models are essential. The traditional modeling methods of spiral inductors such as finite difference time domain method (FDTD), moment method (MOM) and finite element method (FEM) and other numerical methods are cumbersome and time-consuming to calculate. Analytical and empirical modeling techniques can effectively save time, but the calculation The precision is not accurate enough. Additionally, small changes in material parameters may require a simulation rerun. In recent years, some rese...

AI Technical Summary

Problems solved by technology

The introduction of intelligent algorithms into the design of spiral inductors can speed up the design and reduce labor costs, and at the same time design spiral inductors with optimal performance to meet various circuits, which is of great significance to the development of integrated circuits; but it is not enough Meet the needs of circuit design

Images