Maneuvering intelligent decision-making missile avoidance method based on reinforcement learning

The invention discloses a maneuvering intelligent decision-making missile avoidance method based on reinforcement learning, which comprises the following steps of: 1, establishing a reinforcement learning environment at a Unity end; 2, establishing a first neural network which comprises a decision-making network and a curiosity network, and setting an award value judgment mechanism, wherein an award value is composed of a first award value and a second award value, and the first award value is obtained through calculation by using the curiosity network; 3, training the first neural network to obtain a trained neural network in the reinforcement learning environment at a Python end; and 4, utilizing the trained neural network to avoid missiles. The decision-making avoidance neural network is simple in structure, easy to train and good in real-time performance. The maneuvering intelligent decision-making missile avoidance method utilizes the neural network as a mapping function from a state space to an action s