Operational optimisation of smart grid
Smart grid operational optimisation problems are characterised by abundant data, far-reaching control and persuasively
large potential benefits. However, due to time constraints, these problems tend to exceed the tractability limits of exact solution methods. Following an investigation of the trade-off between solution timeliness and accuracy in the context of a receding horizon controller with a heuristic subroutine, this work develops a smart grid controller which employs decision trees and thereby resides at the timely end of the spectrum. The following major contributions have been made:
- Formulation of a theoretical argument for the role of control period selection in the operational performance of smart grids under the simulation-optimisation control scheme.
- Presentation of experimental evidence supporting the identification of control period selection as an important factor in the operational optimisation of smart grids. Development of a new smart grid controller based on a learnt decision tree ensemble.
- Experimental evidence supporting the suitability of the new controller for supply-temperature management in large district heating networks.
- Development and comparison of control strategies for charge scheduling of plug-in hybrid electric vehicles.
KCB Steer, A Wirth, SK Halgamuge, "Decision tree ensembles for online operation of large smart grids", Energy Conversion and Management, 59, 9-18.
Leader: Saman Halgamuge
Staff: Saman Halgamuge,Andrew Wirth
Collaborators: Dr Kent Steer (IBM), Rudolf Kruse (University of Magdeburg)
Networks and data in society, Optimisation of resources and infrastructure