Loading…

Deterministic flood control using MPC on the Kragerø waterways

The Kragerø Waterways consist of 5 power plants downstream after Lake Toke, Dalsfos is the first of these power plants. As a deterministic flood control of Lake Toke and Dalsfos there has been designed a MPC. The MPC uses a horizon of 10 days with estimated influx of water into the Lake Toke water r...

Full description

Saved in:
Bibliographic Details
Main Author: Evensen, Robin
Format: Dissertation
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The Kragerø Waterways consist of 5 power plants downstream after Lake Toke, Dalsfos is the first of these power plants. As a deterministic flood control of Lake Toke and Dalsfos there has been designed a MPC. The MPC uses a horizon of 10 days with estimated influx of water into the Lake Toke water reservoir. As a preliminary task before the MPC is designed, an assessment of the requirements and constraints of Lake Toke was made. The primary constraint is to keep the level between an upper and lower limit regulated by NVE. A 2 state model designed by Bjørn Glemmestad was used to define Lake Toke, to further increase the precision of the model, assessment parameters was done. A simulation and validation of the linearized model of Lake Toke was performed to assess the precision, which proved to perform adequate. A review of the structure of the MPC program is shown along with discussion and explanation of the implementation of constraints and performance requirements. The controller is validated by simulation a flood scenario with disturbances similar to a real flood situation, there was also added a secondary peak to resemble a rainfall. The objective is to keep the level between a lower and upper limit, lrv and hrv, with a goal to avoid an overshoot over hrv. The controller proved to avoid an overshoot when the weight lower limit in the performance index was lower than the upper limit. A weight on the control output also proved to create a more stable control output as well as avoiding an overshoot when the secondary rainfall arrived.