Application of neural networks in process control: Automatic/online tuning of PID controller gains for ±10% disturbance rejection

In the area of process control engineering, PID controller plays a very vital role, which can be used as a process controller for the general industrial processes like liquid level, temperature, pressure, flow, and etc. Most of the research papers in the area of process control deals with setting up...

Full description

Saved in:
Bibliographic Details
Main Authors: Vasu Murthy, B., Pavan Kumar, Y. V., Ratna Kumari, U. V.
Format: Conference Proceeding
Language:English
Subjects:
Algorithm design and analysis
Artificial intelligence
Artificial Neural Networks (ANN)
Intelligent - PID Controller
Liquid level system
MATLAB/Simulink
Open Loop Transient Response (OLTR)
PID Controller
PID Tuning methods
Software packages
Valves
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the area of process control engineering, PID controller plays a very vital role, which can be used as a process controller for the general industrial processes like liquid level, temperature, pressure, flow, and etc. Most of the research papers in the area of process control deals with setting up of these PID controller gains. There were many tuning algorithms evolved starting from the famous Ziegler - Nichols in 1942. But, still it is a major challenge for setting of PID controller gains. Also, since PID controller is an off-line controller, the gains are set in offline and then the controller is put in to the process loop. Hence this conventional off-line PID controller cannot cope up with the online process variations due to the non-linear disturbances/ noises. Hence PID controller is associated with poor disturbance rejection. Based on these typical challenges from the conventional tuning methods, the paper proposes an original philosophy of providing ± 10% disturbance rejection feature to the PID controller. The novelty used in the design is automatic tuning of PID controller gains by the use of intelligent predictors like Artificial Neural Networks (ANN). This provides the facility of online automatic tuning of PID controller gains and so, good disturbance rejection for the system. The whole system is modeled and simulated by using MATLAB/Simulink software. The results show that the system has good disturbance rejection up to ± 10%with the proposed control strategy.
DOI:10.1109/ICACCCT.2012.6320800