Topology and Parameter Estimation in Power Systems Though Inverter Based Broadband Stimulations

During the last decade, a substantial growth in renewable, distributed energy production
has been observed in industrial countries. This phenomenon, coupled with
the adoption of open energy markets has signi cantly complicated the power
ows
on the distribution network, requiring advanced and intelligent system monitoring
in order to optimize the e ciency, quality and reliability of the system.
This thesis proposes a solution several power network challenges encountered with
increasing Distributed Generation (DG) penetration. The three problems that are
addressed are islanding detection, online transmission line parameter identi cation
and system topology identi cation. These tasks are performed by requesting the
DGs to provide ancillary services to the network operator. A novel and intelligent
method has been proposed for reprogramming the DGs Pulse Width Modulator,
requesting each DG to inject a uniquely coded Pseudo-Random Binary Sequence
along with the fundamental.
Islanding detection is obtained by measuring the equivalent Thevenin impedance
at the inverters Point of Common Coupling, while system characterization is obtained
by measuring the induced current frequencies at various locations in the
grid. To process and evaluate the measured signals, a novelWeighed Least-Squares
aggregation method is developed, through which measurements are combined and
correlated in order to obtain an accurate snapshot of the power network parameters.