Abstract: In this paper, we propose an optimal control strategy for electric vehicle aggregator (EVA) participating in secondary frequency regulation ancillary services. This strategy considers the interests of power systems, EVAs and electric vehicle (EV) owners. According to the operating mechanism of frequency regulation market in China’s southern region, a cloud-edge collaboration framework for EVA participating in frequency regulation service is constructed, in which the interactive mechanism of multilevel equipment is clarified in time sequence. In the assessment of EV’s frequency regulation capacity, an edge assessment method for a single EV charging station’s frequency regulation capacity is proposed based on EV real-time status. The numerical results are uploaded to the cloud to predict the scale of EVs’ frequency regulation capacity for declaration in day-ahead market. In terms of the decomposition of commands, a command allocation optimization model is established, with the objective of maximizing EVA’s benefits while taking into account the owners’ needs. The results of the example demonstrate that the strategy is capable of ensuring the effective response of EVA to the commands. It also addresses both the economic requirements of EVA and the charging needs of the EV owners in terms of power allocation and charging time windows.