Abstract: Aiming at the electric machine emulator (EME) system with LCL-type interface, a Kalman observer-based active disturbance rejection dead-beat predictive current control algorithm was proposed to solve the problems of high model dependence, poor robustness, lack of delay compensation consideration and high computational coupling existed in the commonly used current control algorithm, improving its performance by combining Kalman Filter and state observer. Firstly, based on the mathematical model of the LCL-type interface, a predictive control model of dead-beat current was established, considering delay compensation, and the effects caused by the disturbance of the interface parameters and the delay current were analyzed, and the effects caused by the disturbance of the interface parameters were abstracted as the compensation current and the compensation voltage. Further, based on Kalman Filter algorithm, several state observers were designed respectively for compensation current, compensation voltage and delay current. And then, the observed delay current was regarded as the input of the control system, and as the compensation current and compensation voltage to feed back to the control system, achieving precise and stable control of the interface current. The simulation and experimental results show that the proposed Kalman observer-based active disturbance rejection dead-beat prediction control algorithm can effectively improve the system robustness and simulation accuracy.