There is a problem of high climb frequencies of a thermal power unit and large reaction times of a battery energy storage system. Thus an AGC commands tracking control strategy of a thermal power unit assisted by a battery energy storage system based on a swing door trending algorithm is proposed. First, historical data of the AGC are compressed by the swing door trending algorithm and linear interpolation is used to deal with the compressed data. Dynamic update and real-time correction are carried out to obtain the power regulation command of the thermal power unit integrated with its climb feature. Then, the battery energy storage system is dynamically divided into three groups and their action sequences are determined by their power regulation commands. The power distribution approach of the battery energy storage system is designed to realize power allocation in upper-layer and the power allocation principle of the cells for the three packs is devised to accomplish power allocation in the lower-layer. Finally, the simulation of the proposed control strategy is conducted using the AGC commands of an actual power plant and its effectiveness is verified. This work is supported by the Science and Technology Project of China Southern Power Grid Co., Ltd. (No. 036000KK52190005 (GDKJXM20198110)).