Analysis of Tuning Parameters of Model Predictive Controller (MPC)

Process optimization is very important in the engineering industries. As optimisation is achieved, less consumption of energy and utilities can be obtained for the process.
In achieving optimisation, the response should be responded close to the reference values. The refineries nowadays consist mainly of multi variable unit process. Thus, to achieve optimisation using classical approach will be less reliable and time consuming. Hence, the introduction of Model Predictive Controller (MPC) to the
process unit is more suitable compared to the classical approach. MPC is capable to solve high order problem and multivariate processes. The successful of MPC
depends on the selection of tuning parameters. Therefore, by analysing the effect of each tuning parameters on the controller performance, promising performance of MPC can be produced. Firstly, the processes are selected from books as a case study to resemble the high order and multi variable problem processes. Then, the analysis
will be done to study the effect of input weightage (UwO, output weightage (ywt), control horizon (M) and prediction horizon (P) on the controller performance. By
changing one of the tuning parameter, the other tuning parameters have to be kept constant.