Developing and Testing of an MPC strategy for a four tanks multivariable process using Emerson Delta V system.

Abstract

With the advancement of control system, Model Predictive Control has taken over the mainstream PID controller because of its more efficient and advanced control over the multivariable process. The Quadruple Tank process represents the multivariable process, which has 2 input, 2 output and 4 states. The complexity of the control system is handled in a more optimal way with the help of DeltaV MPC system.

The main goal of the thesis is to develop a model that handles the disturbance and the constraints parameter of the Quadruple Tank process. The DeltaV offers a robust controller after successfully developing the model. The controller handles the interaction between the process variables in the best possible way to reach its desired output. The setpoint and the disturbance are handled in a quite satisfactory manner. The step response model for the model developed in the DeltaV system can give a wide operational range, which helps to develop a more robust controller for any situation. After successfully performing the test process, the DeltaV uses the FIR and ARX model to identify the step response model, which will eventually create the optimal controller for the process.

Even though the model showed a satisfactory result but some more development can be achieved by manipulating the instruments such as the pump or the sensor, which will enlarge the operational region as well as handling the disturbances more profoundly.