In this thesis, a fuzzy RBF neural network PID (RBFfnn-pid) control method is designed for temperature deviation concerning the steam curing temperature control system in the steam curing kettle of the aerated concrete block production line. The temperature control system (TCS) models of steam curing are established respectively under the control of fuzzy PID and RBFfnn-pid. Then, the control effect is analyzed by simulation in different situations. The results show that, under this condition of step signal inputting, the RBFfnn-pid control method shortens the rising time by 160 s and reduces the overshoot by 8% compared with the fuzzy PID controlling means. When the system is in a stable state, then the interference signal is added, which in contrast to the controlling means of fuzzy PID, the time for adjusting the RBFfnn-pid control to the set value is shortened by 70 seconds and there is no overshoot. The RBFfnn-pid control method enables the controlling parameters of the steaming temperature control system to be adjusted to the optimum quickly and iteratively, which improves the temperature control accuracy.