Application of Reinforcement Learning in Autonomous Navigation for Virtual Vehicles

Abstract

To solve the basic problem in the autonomous navigation for the virtual vehicles, the application of reinforcement learning is discussed, combined with artificial neural network, a model is proposed to realize the autonomous navigation for virtual vehicles. The autonomous navigation for the virtual vehicles is accomplished by reinforcement learning, the generalization problem for the action selection is solved by neural network, the model is tested by virtual environment, and the experiments show that the model is available and feasible.