Conventional control of tunnel boring machine can't achieve the ideal effect because of the complexity of its work process, uncertainty of construction environment. In order to solve the question, the intelligent control is expected. The neural network control has many characteristics such as self-adaption, self-organization, and can modify the corresponding parameters by learning about external knowledge, so it becomes research focus of tunnel boring machine control system in recent years. In this paper the advantage and disadvantage of several kinds of neural network used in the tunnel boring machine control are introduced, the development trends of tunnel boring machine control is predicted.