Efficient artificial neural network microwave device modeling method based on Bayesian theory

• Main Authors: JIN DONGYUE, XIE HONGYUN, NA WEICONG, LIU KE, ZHANG WANRONG
• Format: Patent
• Language: Chinese; English
• Published: 09.02.2021
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention relates to an efficient artificial neural network microwave device modeling method based on a Bayesian theory, which is used for solving the problems that the existing microwave device neural network modeling method is tedious and time-consuming and is difficult to meet the increasingly improved requirement of shortening the design period of a microwave device and a circuit. According to the method, the optimal effective parameter number required by the corresponding neural network model is firstly found, and then the optimal hidden layer neuron number of the corresponding neuralnetwork model is calculated through the optimal effective parameter number, so that the optimal structure of the neural network model of each microwave device is obtained. Besides, the method can bedirectly embedded into a neural network automatic model generation algorithm, and no matter whether the number of initially set hidden layer neurons is close to the optimal value or not, the neural network model structure opti