The technical performance parameters mentioned in the instruction manual of T8151B electric actuator are not introduced except the external characteristics such as deadband, backlash, hysteresis and operating voltage, and the internal characteristics such as Ton, Toff and Smin. The self-learning function of the PAT module can obtain the characteristic parameters of these electric actuators, which provides a powerful support for the reasonable setting of characteristic parameters.

The design of T8151B self-learning function is based on the principle of artificial neural network and adopts the back propagation algorithm. This design mainly comes from the fact that the characteristic parameters of the electric actuator will change under different loads and different valve positions, and the PAT module determines the control effect of the electric actuator and corrects the characteristic parameters in time according to the control effect, so as to achieve the optimal control.

When T8151B is loaded, the corresponding deadband and threshold are smaller than in the case of no-load, Ton is longer than in the case of no-load, and Toff is shorter than in the case of no-load. In this case, the PAT module can intelligently judge the reasonableness of the set parameters (see Fig. 2, the intelligent processing of Ton pulse is too long is similar, and will not be described in detail) and adjust them in time to ensure that the optimal control is achieved.