Computer Aided Design of Microwave Circuits Based on BP Neural Networks

Abstract

In addition to meeting basic electrical characteristics, the design of a microwave circuit must also consider the ease and accuracy of the circuit model, the time required for circuit design, the availability of tolerance analysis, and the ability to optimize output. Artificial Neural Network (ANN) technology has recently been introduced into microwave circuit CAD and optimization due to its nonlinear mapping function and high real-time computing capability. This technology has been shown to significantly improve the accuracy of modeling and compensate for errors in traditional microwave circuit design methods. In this study, we propose a BP neural network-based modeling method for microwave integrated circuits. This model can be used for computer-aided integrated microsystem simulation and analysis. The fitting error(RMSE) between the original simulation data and the test data is reduced from 0.868 to 0.274 before and after compensation, demonstrating the effectiveness of the proposed method. Experimental results show that the accuracy of the circuit model is improved after the optimization of the BP neural network.