Model Based Algorithms For Quadruple Conical Tank System

Abstract

The Quadruple Conical Tank System (QCTS) is a best example for MIMO process with nonlinear behavior due to its conical shape and with uncountable interactions in it. The objective of the project is to perform numerous model based algorithmic rule for QCTS with simulation study and in real time on QCTS. It is very difficult for classical controllers to control MIMO process with interactions which give way to the implementa- tion of Model based algorithmic rule which involves gathering information from the open loop response data-set and use that as a model. The model based algorithmic design is technique of predefined mathematical model of the system and variations and applying control management on later stage. This model based algorithm involves, modeling of the plant, analyzing the controller, simulation of plant and controller and gathering all the values to develop a model based controller. The approach is to implement Model Predictive Control on Quadruple Conical Tank System and compare its result with PID controller both in simulation and real time. MPC is selected because of its wide variety of application in process control and it uses an explicit model to predict future response of the process. It has the advantage of dealing with more complex and fast dynamics system even in the presence of constraints.