The invention relates to a simulation method in the electric power system scheduling automation field, in particular to a multilayer closed-loop main station detection data source simulation method. The method comprises the following steps of simulating the full-electric-network steady-state real-time data conforming to the operation mechanism of an electric network through mature electric power simulation software, simulating different fault information through a fault generation module, decomposing the fault information to a remote signaling deflection table, an SOE (sequence of event) table, a remote signaling number table, a remote measuring number table and a transmitting cache area which is used for protecting the record in a signal table or directly storing the record in the signal table into a communication module; reading a real-time database according to different periods through the communication module, uniformly organizing the data according to the data priority of number table, deflection table and device table, and finally forming the simulation data formed by combining the variation data and the period data. By adopting the method, the electric network real-time running data of the electric network under the modes such as normal running mode, maximal running mode, disturbance, accident and snow slide can be provided, and a reasonable and more-comprehensive data source is provided for the detection of different tasks of the system.