Synchronous time reference is the key basis of cluster cooperative control.In view of the reduced convergence accuracy and speed of cluster network time synchronization caused by the absence of external precise clock source and the delay of internal communication，a clock synchronization algorithm combining the motion model and state estimation is proposed.In this algorithm，a linear regression model conforming to the clock model is constructed，and the clock compensation parameters are estimated with the pseudo-measurement data.In the self-constructed and maintained cluster unified time base，a time registration algorithm based on two-step weighted fusion is used to achieve the asynchronous measurement data fusion of the target.The simulation results show that the clock synchronization algorithm based on the motion model and state estimation has fast convergence and high clock synchronization accuracy，which can provide a unified time reference for subsequent data fusion，and the time registration algorithm based on two-step weighted fusion has higher accuracy than the one-step Kalman prediction method.