Grid Edge Event Detection Using Multi-Band ML Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Utilities face challenges in managing distributed electrical grids due to issues such as high impedance faults from falling branches, grid stability with distributed energy resources, and the need for rapid event detection and mitigation, as well as effective management of edge equipment.
Innovation Solution
Implementing a sensor system with a fast-throughput, low-latency pattern recognition algorithm and advanced optical current/voltage sensors at the grid edge, integrated with machine learning for real-time event detection and autonomous mitigation, and a cloud-based device management system for remote monitoring and control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If protection systems are used to de-energize conductors with fallen branches, then safety is improved, but response time increases to minutes
Solution Approach 1:
The system performs preliminary action by detecting the presence of fallen branches on conductors before they cause high-impedance faults or wildfires. The optical sensor continuously monitors for branches contacting conductors, and the machine learning algorithm predicts potential fault conditions in advance, allowing the system to de-energize conductors proactively rather than reactively, reducing response time from minutes to seconds while maintaining safety
Solution Approach 2:
The system replaces the traditional mechanical/protection-based detection system with an optical sensing and machine learning-based prediction system. Instead of relying on protection equipment to sense faults after they occur, the patent uses optical sensors to detect branches contacting conductors and ML algorithms to predict impending faults, enabling faster automated response without mechanical intervention
2Ease of operation
If manual management is used for grid edge equipment, then operational control is maintained, but management efficiency decreases and errors increase
Solution Approach 1:
The system enables self-service by allowing grid edge equipment to autonomously detect events, process sensor data, and execute mitigation actions without manual intervention. The machine learning algorithms at the edge automatically predict faults and trigger de-energization commands, while the device management system automatically monitors and manages distributed equipment, eliminating the need for service personnel to visit each site and significantly improving management efficiency
3Measurement precision
If high sampling rate sensor waveforms are used, then detection accuracy is improved, but data processing complexity increases
Solution Approach 1:
The system applies segmentation by dividing the high-sampling-rate sensor waveform data into multiple frequency bands using signal processing techniques. The machine learning algorithm processes each frequency band separately, identifying characteristic patterns that indicate different fault conditions. This segmentation approach maintains high detection accuracy by preserving detailed waveform information while reducing processing complexity through frequency-domain analysis and pattern recognition
Data Source
AI summary
Embodiments detect one or more events on an electrical grid. Embodiments use a sensor installed at an edge of the electrical grid to generate a sensor waveform at a first sampling rate corresponding to current and/or voltage signals. Embodiments transform the sensor waveform into multiple frequency bands and digitize the multiple frequency bands at a second sampling rate that is lower than the first sampling rate. Embodiments receive, by a pattern recognition machine learning algorithm at the edge, the digitized multiple frequency bands for events and predict, using the ML algorithm, an occurrence of the one or more events.