Safety Device and Method for Monitoring an Electrical Energy Supply Network and Computer Program Product
Simple SummaryContent extracted from patent full text and abstract with AI.
The invention describes a safety device and method for monitoring electrical power supply networks using artificial neural networks (ANNs). Unlike traditional protection devices that rely on hard-coded deterministic algorithms, this solution uses deep learning and pattern recognition on graphical representations (like spectrograms) derived from network measurement data. By leveraging image-processing neural networks such as convolutional neural networks (CNNs) and optionally recurrent neural networks (RNNs), the system can more flexibly, quickly, and reliably identify both normal and abnormal operating states, including complex or previously hard-to-detect faults.
Use CasesContent extracted from patent full text and abstract with AI.
- Real-time fault detection in electrical power grids, including subtle fault types like high-impedance ground faults.
- Automated monitoring and protection for decentralized or smart grid systems with renewable energy sources.
- Enhancing substation and transmission line protection schemes.
- Deployment in both local (on-device, edge) and central (cloud-based) grid monitoring infrastructure.
- Smart maintenance systems that analyze grid health and anticipate breakdowns before they escalate.
BenefitsContent extracted from patent full text and abstract with AI.
- Improved detection accuracy, including faults that conventional protective equipment struggles to identify.
- Self-learning and adaptive capability allows the system to adjust to changing grid configurations or operating conditions without extensive reprogramming.
- Faster response to faults, reducing outage times and potential equipment damage.
- Scalable architecture – can be used for single points (local), regional, or grid-wide (central/cloud-based) protection.
- Hardware-agnostic: can be implemented on various platforms, including embedded devices or cloud servers.
- Helps increase operational safety, reduce risk of fires or injuries from undetected electrical faults, and support reliable grid operation.
Technical Classifications (CPCs)
Main Classifications
Electrical & Electronic Tech
Sub Classifications
Electric Power Generation & Distribution
CPC Codes
Inventors & Applicants
Inventors
Applicants
Siemens Ag
Univ Friedrich Alexander Er
Patent Abstract
Die Erfindung betrifft eine Schutzeinrichtung (22) zum Überwachen eines elektrischen Energieversorgungsnetzes (10), mit einer Auswerteinrichtung (40), die dazu eingerichtet ist, unter Verwendung von Messwerten, die einen elektrischen Zustand des Energieversorgungsnetzes (10) an zumindest einer Messstelle angeben, eine Entscheidung zu treffen, ob sich das Energieversorgungsnetz (10) in einem zulässigen oder einem unzulässigen Betriebszustand befindet, wobei die Auswerteinrichtung (40) ein künstliches neuronales Netz umfasst.Um eine besonders flexible und zuverlässige Möglichkeit zur Erkennung unzulässiger Betriebszustände in Energieversorgungsnetzen anzugeben, wird vorgeschlagen, dass die Auswerteinrichtung (40) dazu ausgebildet ist, zur Entscheidung über den Betriebszustand des Energieversorgungsnetzes eine Mustererkennung einer auf den Messwerten basierenden grafischen Darstellung (G) heranzuziehen.Die Erfindung betrifft auch ein mit einer solchen Schutzeinrichtung (22) ausgeführtes Verfahren zur Überwachung eines Energieversorgungsnetzes (10) sowie ein entsprechendes Computerprogrammprodukt.
Key Information
Publication No.
EP4175090A1
Family ID
78414520
Publication Date
2023-05-03
Application No.
EP21205584A
Application Date
2021-10-29
Priority Date
2021-10-29
Granted
No
Possible Cooperation
For further information please contact the transfer office.