A distributed adaptive controller based on radial basis function neural network(RBFNN) is proposed to enhance the transient stability of power systems for the system nonlinearity and parameter uncertainty prevalent in power systems. The unknown nonlinear term and the external disturbance term in the systems are compensated by using the radial basis function neural networks method, and the corresponding adaptive parameter estimation scheme is designed to approximate the ideal weight matrix of the unknown nonlinearity. Based on multi-agent framework, the distributed controller receives real-time data by measuring devices through communication network and controls the action of energy storage device, so that each generator can realize frequency synchronization quickly after disturbance. The convergence of the proposed distributed control method is proved using the Lyapunov stability theory. Finally, simulation studies verify the effectiveness of the proposed distributed control method.