Aiming at the deficiencies of the existing intersection adaptive control methods in terms of the control scheme initiative and the interaction level between the simulation and the actual system, this paper proposes a parallel control method for traffic signals that considers vehicle arrival prediction. The method is based on the theoretical architecture of parallel control, through the construction of a traffic flow arrival prediction model in the artificial system, to achieve the simulation assessment of the adaptability of the signal control scheme in the next time period and the generation of optimization parameters, and then the optimization decision module is designed to generate the best optimization parameters of the actual system. The method is able to improve the perception ability of the actual traffic signal control system and realize the active optimization of the intersection adaptive control scheme through the interaction between the actual system and the artificial system. The experimental results show that this method can effectively reduce the traveling delay and improve the intersection capacity during the time when the traffic flow fluctuates significantly.