(383i) Predictive Control of Lake Levels and Estimation of Stream Flows for the Adaptive Management Complex Natural Watersheds. Part 2. Economic Model Predictive Control
AIChE Annual Meeting
Tuesday, October 31, 2017 - 2:46pm to 3:03pm
This study extends that work by implementing advanced control strategies to control lake levels subject to current rule curves. Model predictive control is a strategy that uses available measurements of Rainy Lake and Namakan Lake levels, a tracking filter to measure ungaged inlet flows, a model of the two reservoir system, and a prediction horizon to estimate an optimal control trajectory. In this instance the control variables are the daily configurations of the dams located at International Falls, Kettle Falls, and the Seine River system. Real-time feedback control is achieved by daily updates of the optimal control trajectory in response to new measurements.
Economic Model Predictive Control (EMPC) is an essential aspect of the proposed control system. The objective incoporates a complex set of water level requirements for key sentinel species including Wild Rice, Walleye and Northern Pike, Loon, and Muskrat.
The primary goal of adaptive management is regulate lake levels and river flows in response to measured events rather than to fixed dates on the calendar. The driving forces behind changes in the hydrology of watersheds in North America is large climate change. In the Rainy River basin, the change in climate has led to earlier snow melt and ice out of the regionâs lakes, and more intense precipitation events in late Spring and early Summer.