Abstract: We propose a distributed adaptive control method to enhance transient stability of power system for transient stability control when partial controllers fail in power system. Based on leader-following consensus algorithm of second order nonlinear multi-agent systems, we obtain the real-time information of neighbor generator through communication network, design a controller based on the information of relative angle and relative speed between the local generator and its neighbors generator, and control the external energy storage system fast recovery transient stability after failure of power system. We prove the convergence by Lyapunov stability theory. Finally, we analyze the performance of the controller in the New England 39-bus test system in depth, and consider the effect of partial controllers failure under cyber attack on the system performance. Simulation studies further validate the feasibility of the proposed distributed control method.