An online BPN gain adapter for PID through PLC on a conical tank system

The PID controller is of great importance to both the researchers and the industrialists as it proves to be robust for linear single input single output processes. However the PID has limitations with non linear and higher order systems. In this research work, an online back propagation neural network gain adapter for PID controller is developed using the ladder functions of PLC and implemented in a non-linear conical tank system. The real time result for steady state performance is compared with the conventional PID controller. The controller track the set point quicker than the PID as the gain values are continuously updated based with respect to the changes in the process. The controller is also validated in real time for servo regulatory and disturbance rejection characteristics on the conical tank system. The online back propagation gain adapter for PID controller developed in PLC is more robust to real time changes and performs better than the PID controller on the conical tank system.