Comparison of Tuning Methods of PID Controllers of Two Conical Tank System of Interacting Type
Abstract
The rapid increasing complexity of modern control systems has accentuated the idea of
applying new modern approaches in order to solve design problems for various control
engineering applications. This paper deals with the tuning of PID controllers for complex
nonlinear process of two interacting Conical Tank Systems. Conical tanks play vital role
in leaching extractions in pharmaceutical and chemical industries, as well as in food
processing industry. It is a kind of multiple-use equipment, designed for boiling and
extraction working procedure. Level control of a conical tank is a tedious process
because of its non-linear characteristics hence proportional-Integral-Derivative (PID)
control schemes have been widely used to overcome the issue by applying different
controlling techniques. A PID controller is otherwise called as three term-control which
has three constant parameters and it takes the present error, accumulation of past errors
and prediction of future errors into account based upon the current rate of change of
error, respectively. The proposed control strategy includes the tuning of PID controller
using Ziegler-Nicholos (Z-N) method and intelligent techniques like Genetic Algorithm
(GA). The scope of this paper is to compare the different conventional tuning methods for
single input single output (SISO) systems such as Z-N, Tyreus-Luben (T-L), Internal
Model Control (IMC) and GA to predict the most efficient controller. The factors for
comparison include time domain specifications and performance of the controller. The
performance of the controller for different tuning rules has been investigated in a
MATLAB simulation Platform in which Genetic Algorithm outperformed well when
compared to all other controllers in terms of time domain specification and performance
index.
Keywords: Conical tank, interacting system, non-linear, PID, controller, performance and genetic algorithm
Keywords
Full Text:
PDFRefbacks
- There are currently no refbacks.