Neuronal adaptive PID control based on recurrent fuzzy neural network for single tank system

In the oil refining industry, chemical industry, water treatment industry, paper production, electricity production... The problem of controlling water level and flow rate needs to be met with high precision to serve the production process and achieve better efficiency. Therefore, the problem is to control the flow rate to stabilize the liquid level with high accuracy. With the requirements of practical application, the article proposes a neural adaptive PID controller based on the recurrent fuzzy neural network identifier (SNA-PID-RFNN) for stable control of liquid flow and online control testing on the water tank system. The simulation results on MATLAB/Simulink show a better response than the traditional PID controller, and the online control experiment on a single tank model to control the stable liquid flow also gives the result that the actual water level follows the reference water level well. The advantage of the SNA-PID-RFNN controller is that it updates and adjusts the controller parameters online, which is suitable for practical applications in industrial production.