Bad data correction in harmonic state estimation for power distribution systems: an approach based on generalised pattern search algorithm

•A novel methodology is proposed for bad data correction in harmonic state estimation;•Bad data is detected, identified and corrected assuming a limited number of measurements;•Three-phase distribution systems are considered for the computational simulations;•Generalized pattern search algorithm is...

Full description

Saved in:
Bibliographic Details
Published in:Electric power systems research 2022-03, Vol.204, p.107684, Article 107684
Main Authors: Melo, Igor D., Antunes, Matheus P.
Format: Article
Language:English
Subjects:
Bad data
Calibration
Distribution systems
Electric power distribution
Electricity distribution
Estimating techniques
Harmonic analysis
Harmonic state estimation
Lower bounds
Mathematical models
Optimization
Pattern search
Power quality
Search algorithms
State estimation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A novel methodology is proposed for bad data correction in harmonic state estimation;•Bad data is detected, identified and corrected assuming a limited number of measurements;•Three-phase distribution systems are considered for the computational simulations;•Generalized pattern search algorithm is used to determine correction factors for bad data. This paper presents a novel methodology for bad data correction in harmonic state estimation for power distribution systems. An optimisation model is formulated considering an objective function to be minimised based on the weighted least squares. Inequality constraints are incorporated to the problem for those buses which are not monitored in real time by any dedicated meter, being their corresponding active and reactive powers considered between upper and lower bounds in order to provide supplementary information about the current system state. In this paper, a measurement calibration vector is introduced into the optimisation model, assuming that a correction factor is associated with the measurements gathered from the network. GPSA (Generalized Pattern Search Algorithm) is used to determine the optimal values of each calibration factor to provide correct state estimation results. A 69-bus test system is used for the computational simulations considering different case studies with multiple bad data to be identified and corrected proving the efficiency and viability of the proposed method. The main contribution of this work is the automatic detection of bad data, identification of the corrupted measurements and correction of the bad data ensuring that the system states are estimated with errors lower than 1%.
ISSN:0378-7796
1873-2046
DOI:10.1016/j.epsr.2021.107684