Electrical Network Diagnosis via Edge Data Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing electrical distribution network failures are inefficient due to long time lags between fault detection and corrective measures, incomplete data accuracy, and the challenge of managing large volumes of sensitive data, which affects data security and resource utilization.
Innovation Solution
A method involving local devices at each network node to record and transmit anomaly data, with a central device processing and enriching this data to identify relevant events, distinguish between network modifications, and simulate interventions for precise diagnostics and maintenance planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive real-time monitoring data is collected and centralized from all network nodes, then diagnostic accuracy and coverage are improved, but data security risks and resource consumption increase
Solution Approach 1:
The patent segments the centralized monitoring system into distributed edge computing nodes. Each edge device processes and filters data locally before transmitting only essential information to the central system. This segmentation reduces the volume of sensitive data that needs to be centralized and transmitted, thereby maintaining diagnostic accuracy while reducing data security risks associated with centralization.
Solution Approach 2:
The patent introduces edge computing devices as intermediaries between the network nodes and the central monitoring system. These intermediary devices perform preliminary data processing, filtering, and anonymization, so that only aggregated and non-sensitive information is transmitted to the central system. This intermediary layer preserves diagnostic value while protecting sensitive customer data.
2Measurement precision
If detailed data is collected from all network nodes, then diagnostic precision is improved, but bandwidth consumption and processing resources increase
Solution Approach 1:
The patent extracts and processes data at the edge devices before transmission to the central system. Each edge device performs local filtering, aggregation, and preprocessing of raw data, extracting only the most relevant features and anomalies. This extraction approach maintains diagnostic precision by preserving critical information while dramatically reducing the volume of data that consumes bandwidth and processing resources during transmission and centralized processing.
Solution Approach 2:
The patent implements partial data collection by having edge devices selectively transmit only anomaly-related data and aggregated statistics rather than all raw data. This partial action approach focuses computational and communication resources on the most diagnostically valuable information, maintaining precision where needed while reducing overall resource consumption.
3Reliability
If periodic simulations are performed on theoretical network topology, then potential failures can be predicted, but the time elapsed before corrective measures is significant
Solution Approach 1:
The patent performs preliminary failure detection and diagnosis continuously at the edge devices using real-time data processing. Instead of waiting for periodic simulations to identify issues, the system proactively detects anomalies as they occur and immediately initiates diagnostic workflows. This preliminary action at the edge significantly reduces the time between fault occurrence and corrective measures while maintaining reliability through continuous monitoring.
Solution Approach 2:
The patent implements preliminary anti-action by having the system automatically respond to detected anomalies with pre-planned corrective actions. When edge devices detect specific anomaly patterns, they trigger automated responses such as isolating affected segments, rerouting power flow, or alerting maintenance teams before failures propagate. This preliminary anti-action reduces response time while maintaining system reliability.
4Device complexity
If human operators manually monitor and diagnose network issues, then diagnostic accuracy can be maintained with limited data, but the time elapsed before corrective measures remains significant
Solution Approach 1:
The patent implements self-service through automated diagnostic algorithms deployed at edge devices. These algorithms independently analyze local data, identify anomalies, and generate preliminary diagnoses without requiring continuous human intervention. The system serves itself by performing real-time self-diagnosis and automatically initiating corrective workflows, which dramatically reduces the time from anomaly detection to corrective action while maintaining diagnostic quality through sophisticated automated analysis.
Solution Approach 2:
The patent establishes continuous feedback loops where edge devices monitor network conditions, automatically adjust operations based on detected anomalies, and learn from outcomes. The system uses feedback from real-time data and automated diagnostic results to continuously improve its detection and response capabilities, reducing response times while maintaining or improving diagnostic accuracy through adaptive learning.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
A method for diagnosing an electrical network comprising: a. recording (101) events (21, 22, 23, 24, 25, 26) on a local device, each event being defined at least by an identifier of the node and/or local device, a date and a value quantifying an anomaly in at least one measured property; b. performing a general collection (102) of the records on a central device (10) connected to the network; c. supplementing (103) each collected event with network tree data; d. comparing (104) the events with historical data; e. ordering (105) the events with respect to each other; f. selecting (106) a subset of the events (22, 25) according to the ordering; g. enriching the selected data by performing a targeted collection of additional data.