Chiral meta-structure surface structure parameter optimization model training method, intelligent agent and device

The invention provides a chiral super-structure surface structure parameter optimization model training method, an intelligent agent and a device, and the method comprises the steps: training a chiral super-structure surface structure parameter optimization model based on a near-end strategy optimization (PPO) algorithm in a reinforcement learning algorithm, and transmitting an output action of the structure parameter optimization model to an environment in a Markov decision system, enabling the forward prediction neural network in the environment to output a spectral response, calculating an award according to the spectral response, and receiving a next observation state and a new award to iteratively train the structure parameter optimization model again. According to the method and the device, the response speed of the interaction environment can be effectively improved on the basis of ensuring the optimization accuracy of the structure parameters of the chiral super-structure surface, so that the efficien