Different-gluten wheat variety analysis system based on reinforcement learning

The invention relates to the technical field of reinforcement learning, and particularly discloses a different-gluten wheat variety analysis system based on reinforcement learning, which comprises a different-gluten wheat variety and wheat flour phenotype associated data module, a different-gluten wheat variety reinforcement learning module, a wheat purity measurement module and a wheat flour product taste and flavor change evaluation module. According to the method, the gluten change coefficient is introduced into the reward unit of reinforcement learning, so that the analysis error of wheat varieties with different gluten degrees in the reward unit of the reinforcement learning module can be reduced, and real-time dynamic data of the wheat hardness, the wheat flour protein content, the cereal gluten protein content and the chain protein content can be quickly obtained; according to the method, the wheat environment content in reinforcement learning training can be enriched, the purity of wheat with differen