Model Predictive Control with Shell Standard Control Problem

A DISSERTATION OF CHEMICAL ENGINEERING

Design Model Predictive Control on Shell Control Problem

There are so many researches about Shell Control Problem. This said, the process is strongly interactive and there are a great number of input as well as outputs. PID controller could do the good job for controlling the process but Model Predictive Controller could definitely do better.

Solution to the Shell standard control problem using genetically tuned PID controllers\" (as attached) contained the previous research about how to using genetic PID control algorithms to control the \"difficult process\" (Shell control problem). Hence, it include the process model (in appendices).

You need to implement this in SIMULINK (MATLAB) with PID controller as well as process model. Do as the same in the paper. Provide pictures of simulink diagram, the graph of how process output successfully approach setpoint as the paper has shown, and the matlab code (if applicable) etc. This is design/simulation work

- After that, replace PID controller by Model Predictive Control (MPC) by using MODEL PREDICTIVE CONTROL TOOLBOX (MATLAB). Change prediction horizon, control horizon, weight, etc. to witness the effectiveness of MPC. Also provide simulink diagram, the graph and matlab code (if applicable) This is also the design/simulation work.

- The dissertation focus on MPC so it should be explained why MPC is superior to other control algorithm (PID control algorithm). A comparison should be include to demonstrate the point of view. You need to compare outcome graph that the process output for MPC also approach setpoint but it is not oscillatory and is more stable comparing to PID controller, ect