In industrial production, liquid level control system has been widely used, but for this complex control system with large delay and non-linearity, the traditional PID control method has the drawbacks of difficult parameter tuning and unsatisfactory control effect. Based on the research of traditional PID algorithm, fuzzy control algorithm and neural network algorithm, this paper puts forward a solution of applying fuzzy neural network PID algorithm to liquid level control system. Matlab is used to simulate the control of liquid level object, and A3000 water tank experimental platform is used to verify the simulation results. The results show that the liquid level control system based on the PID algorithm of the fuzzy neural network is superior to the traditional PID algorithm in adjusting time and overshoot, and has better control effect and anti-interference ability, which overcomes the shortcomings of the traditional PID algorithm.