Power-Line Event Location Using Unsupervised Feeder Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Power-line event location identification in distribution feeders is challenging due to the limited number of measurement devices and the unavailability of network parameters.
Innovation Solution
An unsupervised approach using synchro-waveform measurements from line-mounted sensors to identify event locations without requiring prior knowledge of network parameters, involving approximation of voltage waveforms, estimation of network parameters, and reconstruction of a feeder model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If synchro-waveform measurements from line-mounted sensors are used, then event location identification accuracy is improved, but the requirement for network parameters knowledge increases
Solution Approach 1:
The system uses the synchro-waveform measurements themselves to extract and estimate the network parameters (line impedance, loading conditions) that are normally required as prior knowledge. The measurements serve dual purposes: both for event detection and for characterizing the network, eliminating the need for separate parameter acquisition systems.
Solution Approach 2:
The invention transforms the approach by changing from requiring fixed network parameters as input to dynamically estimating these parameters from the waveform measurements themselves. This parameter transformation allows the system to operate without pre-existing network models while maintaining high location accuracy.
2Loss of information
If traditional measurement devices are used, then network parameters can be obtained, but the device complexity and installation cost increase
Solution Approach 1:
The line-mounted sensors are designed to perform multiple functions: they measure synchro-waveforms for event detection, estimate network parameters, and provide location identification. This multi-functionality replaces the need for separate specialized measurement devices that would otherwise be required to obtain network parameters.
Solution Approach 2:
The invention replaces complex physical measurement systems with a signal-processing-based approach. Instead of using dedicated devices to measure network parameters directly, the system uses waveform analysis and signal processing algorithms to extract parameter information, reducing hardware complexity.
3Measurement precision
If comprehensive network models are used, then event location accuracy is improved, but the system setup time and complexity increase
Solution Approach 1:
The system performs preliminary estimation of network parameters directly from the synchro-waveform measurements during the event detection process itself, rather than requiring pre-configured network models. This eliminates time-consuming setup procedures while maintaining the accuracy benefits of using network information.
Data Source
AI summary
Disclosed are various embodiments for unsupervised power-line event location within a power-line network. In one embodiment, e-field measurements are received from a plurality of line-mounted sensors located on one or more power lines in the power-line network. The e-field measurements to voltage approximations. A feeder model of the power line network is constructed based at least in part on the voltage approximations. An event within the power-line network is identified. A location of the event is identified based at least in part on the feeder model.