The invention discloses a hypersonic flight vehicle control method based on parameter learning finite-time convergence and aims to solve the technical problem of poor practicability of an existing hypersonic flight vehicle control method. According to the technical scheme, modeling error information is obtained by constructing a first-order filter and an auxiliary signal, an RBF neural network method is designed for unknown nonlinear function estimation in a hypersonic flight vehicle model by combining with self-adaption laws of tracking error and auxiliary signal design neural network, and the thought of finite time learning is introduced to neural network weight updating law design. Compared with a traditional neural network method, the modeling error information is introduced to weight updating laws by constructing the first-order filter and the auxiliary signal, parameter learning finite-time convergence can be ensured, and accordingly the learning rapidity is ensured. Due to the fact that uncertainty is estimated by adopting neural network learning without linear parameterization expression of the model, uncertain control to the hypersonic flight vehicle can be achieved, and practical engineering application is promoted.