Automatically Detecting and Correcting Errors in Power Quality Monitoring Data

Abstract

Abstract-Dependable power quality (PQ) monitoring is crucial for evaluating the impact of smart grid developments. Monitoring schemes may need to cover a relatively large network area, yet must be conducted in a cost-effective manner. Real-time communications may not be available to observe the status of a monitoring scheme or to provide time synchronization, and therefore undetected errors may be present in the data collected. This paper describes a process for automatically detecting and correcting errors in PQ monitoring data, which has been applied in an actual smart grid project. It is demonstrated how to: unambiguously recover from various device installation errors; enforce time synchronization between multiple monitoring devices and other events by correlation of measured frequency trends; and efficiently visualize PQ data without causing visual distortion, even when some data values are missing. This process is designed to be applied retrospectively to maximize the useful data obtained from a network PQ monitoring scheme, before quantitative analysis is performed. This work therefore ensures that insights gained from the analysis of the data-and subsequent network operation or planning decisions-are also valid. A case study of a UK smart grid project, involving wide-scale distribution system PQ monitoring, demonstrates the effectiveness of these contributions. All source code used for the paper is available for reuse.

Data‐driven partition control strategy for harmonic distortion and voltage unbalance

“…In a real distribution network, the data of THD V and TVU are measured using PQ monitoring devices at buses. The measured data in the entire distribution network are collected and stored by the information management system of smart grids [22]. UDTS and MDTS can be constructed from a year of measured data, which reflect secular variations in single and integrated disturbances.…”

Topology-Based Estimation of Missing Smart Meter Readings

“…14:00 and from 14:30 to 15:30. The proposed method is compared with two popular methods: the NNbased method presented in [15,20] and the average method adapted in [12,13]. …”

“…Approaches to compensate for these missing readings included the interpolation of the missing readings by taking an average of adjacent correct readings [12,13]. The results from this approach hardly convinced customers to pay their bills because the mean value is not a good representation of the actual usage.…”

Self‐adaptive learning based controller to mitigate PQ issues in internet of things devices

“…Any distortion in main distribution current is the resultant of phasor sum of multiple harmonic current components, ignited by the non-linearity. 12 Here, D is the distortion component and affected directly by the respective orders of currents…”