This thesis develops solutions that accommodate the introduction of state estimation in High Voltage (HV) power distribution networks, and proposes methodologies that further enhance the value of state estimation in distribution network operation. Furthermore, it investigates the impact of Distributed Generation (DG) and Active Management (AM) on the infrastructure development of typical European distribution networks. In HV distribution networks, measurements are very limited and normally available at the main substation only. Thus, it is essential to introduce appropriately modelled pseudo measurements. This is necessary not only for the state estimation mathematical models to be established but also for state estimation to generate estimates of sufficient quality. Two approaches, one based on correlation coefficients and regression analysis and one based on Artificial Neural Networks (ANNs), are proposed. Distribution networks are not static. Faults, maintenance and emergencies constantly change their topology; sudden changes in major power injections significantly change their power flows and voltages. For the Distribution Management System (DMS) to be reliable, it is important that changes significantly changing the state of the network are immediately identified and taken into consideration before control actions are issued. A methodology for detection of network changes using state estimation and the Bayes theorem of conditional probability is introduced. Finally, the future infrastructure development of typical distribution networks of Germany, the Netherlands and Poland is examined. The technical, economic and environmental aspects of Passive Management (PM), AM and DG are assessed and quantified while the technical and economic efficiency of different AM strategies is evaluated.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:550819 |
| Date | January 2012 |
| Creators | Manitsas, Efthymios |
| Publisher | Imperial College London |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0016 seconds