The invention provides an overload control method for an electric power communication network application server based on hybrid optimization, and belongs to the technical field of electric power communication networks. According to the overload control method, an energy efficient optimal overload control model is built by setting a relation function between a control rate and energy efficiency and using overload control as a constraint condition and through the consideration of the problems of overload control and energy efficiency of the application server. The overload control is carried out by using hybrid particle swarm optimization and a genetic algorithm, the search efficiency of the hybrid particle swarm optimization is high, so the overload control can be limited to a range near to an extremal solution through less algebra evolution. Update limitation is carried out on particles by using a simulated annealing strategy so that the particles can jump out of the local optimum, enable a searching scope to be broadened and get rid of a prematurity phenomenon. Finally, individual assessment is carried out by using an applicable degree function, and genetic manipulation is carried out on this basis to obtain an accurate optimal solution. The overload control method can effectively carry out overload control on the application server in the electric power communication network.