Dual-mode coexistence photonic crystal nano beam cavity reverse design method based on neural network

The invention belongs to the technical field of photonic crystal nano beams, and particularly relates to a reverse design method of a dual-mode coexistence nano beam cavity based on a neural network. According to the method, the structural design of the dual-mode coexistence nano-beam cavity according to the target transmission spectrum is realized; and the design is divided into two stages of training a reverse neural network model and calling a reverse model to carry out structural design. Firstly, in a training stage, data for training is acquired through a finite difference time domain method; then establishing a full-connection forward network with a structure as input and a transmission spectrum as output; the forward network and a network with a transmission spectrum as input and a structure as output are spliced to train a reverse network; and after the reverse network is trained, the target transmission spectrum meeting the dual-mode characteristics can be used as the input of the reverse model in th