Benchmarking the operation of a hydro power network through the application of agent-based model predictive controllers

Reference:
J.M. Maestre, M.D. Doan, D. Muñoz de la Peña, P.J. van Overloop, T. Keviczky, M.A. Ridao, and B. De Schutter, "Benchmarking the operation of a hydro power network through the application of agent-based model predictive controllers," Proceedings of the 10th International Conference on Hydroinformatics (HIC 2012), Hamburg, Germany, 8 pp., July 2012. Paper HIC2012_0338.

Abstract:
This paper presents a comparison between a decentralized and a distributed model-based predictive controller on a hydro power valley (HPV) benchmark recently proposed. The HPV is composed by three lakes and a river that is divided in six reaches that terminate with dams equipped with turbines for power production. The lakes and the river reaches are connected in three different ways: by a duct, ducts equipped with a turbine, and ducts equipped with a pump and a turbine. The river is fed by upstream inflows and tributary flows. In order to test the controllers, the following test scenario is considered: the power output of the system should follow a given reference while keeping the water levels in the lakes and at the dams as constant as possible. Finally, a 24 hour simulation for the two controllers has been carried out in order to compare both methods using several performance indices.