In order to meet the demand of tension stability in reel-to-reel multifunctional printing machine system, the work aims to propose a new tension control method to solve the tension coupling problem and enhance the anti-interference ability and robustness of the system. By analyzing the operating mechanism of the tension system, the two-stage tension coupling mathematical model between the unwinding unit and the printing unit was established, and the dynamic decoupling and static decoupling models of the system were derived. Based on the Active Disturbance Rejection Control (ADRC) method, Radial Basis Function (RBF) neural network was introduced to adjust the key parameters of the nonlinear composite part online. An RBF neural network active disturbance rejection controller (RBF-ADRC) was designed. Compared with the simulation results of other controllers, RBF-ADRC could effectively realize the decoupling control of the tension system, significantly improve the constant tension stability of the system, and enhance the internal robustness and anti-interference ability. RBF neural network active disturbance rejection control strategy successfully solves the tension coupling control problem, which not only realizes the stable operation of the system at constant tension, but also significantly improves the anti-interference ability and robustness of the system, achieving the research goal.