Dynamic Network Island Event Localization by Component
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current predictive maintenance applications for complex networks, such as electrical grids, struggle to accurately identify the specific components causing interruptions, leading to inefficient response and control, especially during dynamic changes like islanding and remerging events.
Innovation Solution
A machine learning latent variable model is employed to disaggregate network events, allowing for the precise assignment of interruptions to individual components, using probabilistic modeling and iterative probability updates to predict future network behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simple logic is used for predictive maintenance, then interruption events can be predicted for relatively simple networks, but the system cannot accurately identify specific components causing interruptions in complex networks
Solution Approach 1:
The patent segments the complex network into multiple islands based on topological connectivity. Each island represents a distinct operational zone that can be independently analyzed. This segmentation allows the system to apply predictive maintenance logic to manageable subsets of the network rather than attempting to analyze the entire complex network at once, thereby improving component identification accuracy without requiring the system to handle full network complexity simultaneously.
Solution Approach 2:
The patent introduces a new dimension of analysis by incorporating island membership and topological relationships into the predictive maintenance framework. Instead of analyzing components in isolation or using traditional single-dimension approaches, the system evaluates events within the context of island structures and connectivity patterns, enabling more precise component identification in complex networks by adding spatial and topological dimensions to the analysis.
2Quantity of substance
If network complexity increases, then more components can be monitored, but the logic of predictive maintenance becomes much more difficult
Solution Approach 1:
The patent divides the large-scale network into multiple islands, each containing a subset of monitored components. This segmentation allows the predictive maintenance logic to be applied to smaller groups of components within each island rather than to the entire network simultaneously, reducing the computational and logical complexity while maintaining monitoring of all components across the full network.
Solution Approach 2:
The patent implements dynamic island formation and dissolution based on real-time network conditions and event patterns. As network conditions change, islands are dynamically reconfigured to optimize the application of predictive maintenance logic. This dynamic approach allows the system to adapt to varying network complexities and component quantities, maintaining manageable logic complexity even as the number of monitored components increases.
3Measurement precision
If traditional predictive maintenance is used, then broad region assignments can be made, but specific components cannot be pinpointed during dynamic islanding events
Solution Approach 1:
The patent segments the network into islands that dynamically form and dissolve based on event patterns. During islanding events, this segmentation allows the system to precisely localize events to specific islands and their constituent components rather than assigning them to broad regions. The segmented structure provides the granularity needed for accurate event localization while maintaining the adaptability to handle dynamic network reconfigurations.
Solution Approach 2:
The patent employs dynamic island structures that automatically adjust to network conditions during islanding events. This dynamic capability allows the system to maintain accurate event localization precision even as the network topology changes, because the island assignments are continuously updated to reflect current network states rather than relying on static regional definitions.
Data Source
AI summary
The present disclosure provides systems and methods that may advantageously apply machine learning to detect and ascribe network interruptions to specific components or nodes within the network. In an aspect, the present disclosure provides a computer-implemented method comprising: mapping a network comprising a plurality of islands that are capable of dynamically changing by splitting and/or merging of one or more islands, wherein the plurality of islands comprises a plurality of individual components; and detecting and localizing one or more local events at an individual component level as well as at an island level using a disaggregation model.