Performance Analysis of Process Using Model Predictive Controller Strategy

Abstract

Distillation is a method of separating mixtures based on different in their boiling points. It is a
physical separation and not a chemical operation. McCabe-Thiele method used to find the no of
trays in a distillation column of tray type distillation column. In this thesis we will see how and
why certain variables may be manipulated to control product composition in distillation column,
to control distillation column we used model predictive control.
Model predictive control (MPC) has become the leading form of advanced multi variable control
technique in process industry. With the help of this thesis we want to presents reliable tuning
strategy for unconstrained single input single out put (SISO) dynamic matrix control (DMC).
The tuning strategy achieves set point with minimal over shoot and modest manipulate input
move sizes and it applicable to a broad class of open loop stable process.
We used DMC algorithms for model control algorithm; in this thesis explain Single input Single
output and Multi input and Multi output DMC algorithms.
This thesis presents a model predictive control strategy for multivariable nonlinear control
problems2 2,3 3 and 4 4process in distillation column,also explain how the tuning
parameters affect the step response model of water heater.
The aim is to provide a solution to nonlinear control problem that is favorable in terms of
industrial implementation. MPC TOOLS of MATLAB® has used to simulate the all process.