To solve the conflict of energy-saving, time-saving and comfort better, the train operation optimization control model is built under many constraints, and the condition conversion points are the goals for the model. And Optimal Operation of the Bend (OOB) is proposed as the rule for train, the operation strategy for the corners is optimized. By analyzing the model, Genetic Algorithm (GA) is used to solve this optimization model. In the end, a simulation is made to validate the model with MATLAB. From the case analysis, it proves the effectiveness of the given method to optimize the operation model, which reduces the energy consumption by 33.27% with only 11.96% extended at run time. At last, the good operation strategy is given.