This paper focuses on train running speed tracking control of electric traction freight train. The physical model of freight train is more complex and the parameters are more unstable than other trains. Therefore, the traditional PID controller is affected by the unmodeled dynamic and unknown parameter changes, resulting in large and unstable train speed tracking error. The design of NLADRC(Nonlinear Active Disturbance Rejection Controller, NLADRC) algorithm based on the time-delay control model of train in this paper. It reduces the dependence on the train model itself and has good tracking performance. The NLADRC gives full play to the strong function and high efficiency of the nonlinear function, while the introduction of the artificial bee colony algorithm with excellent experience effectively solves the problem that the parameters of the NLADRC algorithm are difficult to be adjusted. In the simulation, part of the mechanical delay of the system are regarded as internal and external disturbance, the results show that compared with the traditional PID algorithm, this algorithm has the advantages of high tracking accuracy, strong anti-interference ability and faster response speed.