Model development for real time optimal control in pipe lines

A model reference optimal control architecture for the real-time fluid control of eliminating a contaminant plume from a pipe system was introduced in an earlier paper [1]. The mathematical model for the contaminant flow, is now extended and parameterized based on computational fluid dynamic simulations of a two-dimensional (2D) channel. It is also shown, based on the computational fluid dynamics (CFD) simulations, that the 2D-based mathematical model can be used for three-dimensional (3D) pipe flow problem under certain constraints. Finally, we also derive a very simple control law of the flow rate in a boundary port that would remove a contaminant if the flow was a theoretical 2D flow. This simple control law can initialize the iterative process of computing the optimal flow based on a more complex model and real-time measurements. With the results in this article, the optimal control architecture can be tested in the real-time prototype experiments. Further improvements on the mathematical model and control algorithm can be made for real life pipe-line fluid control problems.