Abstract: Aiming at the problems of nonlinear, strong coupling and multivariable disturbance in the pressure control system of vacuum suction casting equipment, the vacuum suction casting process was analyzed, and a fuzzy neural network PID controller for real-time pressure optimization was designed. Using BP neural network to strengthen the on-line reasoning function of fuzzy algorithm, a nonlinear model between casting quality, minimum thin-walled thickness, material density,pressure of suction chamber and PID correction was established, and then the PID output was adjusted on-line by fuzzy algorithm. The Matlab software was used for simulation analysis and compared with PID and fuzzy PID. The results show that the fuzzy neural network PID has no overshoot, which can not only make the pressure meet the process setting, but also effectively control the decompression speed and other parameters. It has good robustness and meets the requirements of pressure control in the process of filling and solidification.