Automated Electrical Network Mapping via Graph Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current electrical distribution network management relies heavily on manual updates of Geographic Information Systems (GIS) data, leading to inaccuracies and inefficiencies, and lacks standardized compatibility for information exchange among actors, resulting in complex and costly data maintenance and potential incidents due to errors.
Innovation Solution
A computer-implemented method for automating the construction, correction, and updating of electrical distribution network maps using existing communicating devices, such as smart meters, which count communication routes between nodes and apply graph partitioning algorithms to identify and correct errors without human intervention, utilizing Powerline Communication architecture to facilitate data exchange.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network mapping is performed frequently to ensure up-to-date information, then mapping accuracy is improved, but network traffic overhead and system resource consumption increase
Solution Approach 1:
The patent implements dynamic network mapping by transitioning from static periodic updates to event-driven updates. The mapping information is updated only when specific events occur (link up/down, IP address changes, VLAN changes), making the system adaptive to actual network changes rather than following a fixed schedule. This resolves the contradiction by updating mappings dynamically based on need rather than frequency.
Solution Approach 2:
The patent enables network devices to autonomously detect and report changes in network topology and parameters. Devices perform self-diagnosis and automatically trigger mapping updates when they detect changes, eliminating the need for centralized polling or frequent scheduled updates. This self-service mechanism ensures mapping accuracy while minimizing unnecessary resource consumption.
2Loss of information
If detailed network mapping information is collected to improve monitoring capability, then network visibility is improved, but information security risks and system complexity increase
Solution Approach 1:
The patent segments network mapping information by functional categories (physical topology, logical topology, service mappings). Different types of mapping information are organized separately and can be managed independently. This segmentation reduces system complexity by breaking down the monolithic mapping system into manageable modules while maintaining comprehensive network visibility through the aggregation of segmented information.
Solution Approach 2:
The patent introduces a standardized mapping information model as an intermediary layer between network devices and the monitoring system. This intermediate representation layer simplifies the interface complexity by providing a unified, standardized format for diverse network information, while preserving detailed network visibility through the comprehensive data captured in the standardized model.
3Reliability
If network mapping updates are performed in real-time to ensure accuracy, then mapping reliability is improved, but network traffic overhead increases
Solution Approach 1:
The patent implements periodic validation of mapping information combined with event-triggered updates. Instead of continuous real-time updates, the system performs periodic consistency checks and only initiates full updates when validation fails or specific events occur. This periodic action maintains mapping reliability while significantly reducing network traffic overhead compared to continuous real-time updates.
Solution Approach 2:
The patent extracts and processes only the essential mapping information needed for operational reliability, filtering out redundant data. By taking out only the critical mapping elements that directly impact network operation and reliability, the system reduces the volume of data that needs to be transmitted and processed, thereby reducing network traffic overhead while maintaining necessary reliability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for constructing a mapping (100) of an electrical distribution network comprising: a) during at least one observation period and for each pair of consumer nodes, count the number of times that a corresponding link forms part of a communication route, b) construct a graph in which the weight of each arc between two nodes is allotted as a function of the number counted, c) partition the vertices of the graph into groups by applying algorithms devised so that: - two vertices between which the weight of the arc is high and/or two vertices sharing a high number of common neighbours tend to be classed in one and the same group, whilst - two vertices classed in distinct groups tend to exhibit a low or zero link weight and/or to possess few common neighbours.