The invention provides a unit start and stop optimization method with electric automobile charging being considered and based on an ant colony algorithm. The method is characterized by selecting costas an objective function, and establishing a unit combination model with electric automobile charging being considered and a plurality of constraint conditions being considered; carrying out unit start and stop optimization by utilizing the ant colony algorithm to obtain optimal unit combination and minimum power generation cost; and making an output plan of units of an electrical power system anda charge and discharge control plan of the electric automobiles at a special time. Different from unit combination without electric vehicles, the method, considering large-scale electric vehicle networking, carries out joint optimization on traditional unit combination and V2G technology, and utilizes stored energy of a large number of electric vehicles as a buffer for a grid and renewable energy, so that the method plays a filling valley and modulating peak role for the grid, reduces unnecessary energy waste and operation cost of the power grid, is of great significance to operation and dispatch of the power grid and improves networking technique of the electric automobiles and unit combination practicality of the electric power system.