Virtual Metering With Dynamic Tolerance Curves for Power Quality Events
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Power quality issues in electrical systems are costly and prevalent, with 80% of disturbances occurring within facilities, causing significant economic losses and operational disruptions, and existing technologies lack effective methods for characterizing and mitigating these issues.
Innovation Solution
The system and method involve processing electrical measurement data from intelligent electronic devices (IEDs) to generate dynamic tolerance curves, which characterize power quality events and their impacts on electrical systems, allowing for real-time monitoring and adjustment of alarm thresholds, and identifying recovery times and costs associated with power quality events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional power quality monitoring methods are used, then basic power quality parameters can be measured, but the system cannot effectively characterize or mitigate power quality events
Solution Approach 1:
The patent replaces traditional mechanical/electrical power quality monitoring methods with data science and machine learning approaches. Specifically, it uses supervised learning algorithms to train power quality event classifiers, transforming the monitoring system from reactive measurement to predictive characterization, thereby enabling effective power quality event identification and mitigation strategy development
Solution Approach 2:
The patent transforms power quality monitoring by changing the parameters being analyzed from basic electrical measurements to comprehensive feature sets including voltage, current, power, frequency, and derived features. This parameter expansion enables more precise characterization of power quality events and supports the development of targeted mitigation strategies
2Measurement precision
If dynamic tolerance curves are generated for multiple metering points, then power quality event characterization improves, but system complexity increases
Solution Approach 1:
The patent implements a virtual power quality event classifier that serves multiple metering points simultaneously. This universal classifier processes data from various locations through a centralized machine learning model, eliminating the need for separate classifiers at each metering point while maintaining accurate power quality event characterization across the entire electrical system
Solution Approach 2:
The patent introduces a virtual power quality event classifier as an intermediary component that bridges multiple metering points and the power quality monitoring system. This virtual classifier acts as a mediator that consolidates data from various locations and provides unified power quality event characterization, simplifying the overall system architecture
3Reliability
If recovery time and cost analysis is implemented, then operational resilience improves, but data processing requirements increase
Solution Approach 1:
The patent implements preliminary action by pre-training power quality event classifiers with historical data and pre-establishing mitigation strategies for different event types. When power quality events occur, the system can immediately retrieve and apply pre-determined mitigation actions, reducing real-time data processing requirements while maintaining high operational resilience
Solution Approach 2:
The patent enables self-service through automated mitigation strategies that execute without extensive human intervention or complex real-time analysis. The system automatically identifies power quality events, retrieves appropriate mitigation strategies from pre-established databases, and implements corrections, thereby reducing data processing energy requirements while improving operational resilience
Data Source
Figure 1
Figure 1A
Figure 1B
AI summary
A method for characterizing power quality events in an electrical system includes deriving electrical measurement data for at least one first virtual meter in an electrical system from (a) electrical measurement data from or derived from energy-related signals captured by at least one first IED in the electrical system, and (b) electrical measurement data from or derived from energy-related signals captured by at least one second IED in the electrical system. In embodiments, the at least one first lED is installed at a first metering point in the electrical system, the at least one second lED is installed at a second metering point in the electrical system, and the at least one first virtual meter is derived or located at a third metering point in the electrical system. The derived electrical measurement data may be used to generate or update a dynamic tolerance curve associated with the third metering point.