Safety Device and Method for Monitoring an Electrical Energy Supply Network

Simple SummaryContent extracted from patent full text and abstract with AI.

This patent describes a protection device and method for monitoring electrical power supply networks using a single, specially trained neural network instead of traditional logic or multiple algorithms. The neural network receives measurement data from the network, processes it to execute all necessary protection functions in parallel (such as fault detection, fault type classification, and location), and makes fast, adaptive decisions about whether the network is operating normally or abnormally. The device can be deployed either locally (decentralized) or centrally, and is easily re-trainable to adapt to changes in network topology or operational conditions.

Use CasesContent extracted from patent full text and abstract with AI.

• Real-time monitoring and protection of power grids against faults such as short circuits and earth faults.
• Enhanced grid security in smart grids with distributed energy resources (e.g., renewable energy installations).
• Adaptive protection for utility substations or industrial power distribution systems.
• Implementation in microgrids to support autonomous fault detection and response.
• Integration into cloud-based or modern digital grid management platforms.

BenefitsContent extracted from patent full text and abstract with AI.

• Enables much faster and more adaptive decision-making compared to traditional protection systems, thanks to the neural network's parallel processing.
• Capable of adapting to changing grid conditions without the need for complete hardware or software overhauls.
• Reduces reliance on deterministic and fixed algorithms, increasing flexibility and resilience.
• Supports both decentralized (on-site) and centralized (cloud or control center) deployment, providing scalability.
• Improves fault localization and classification, leading to quicker and more precise grid reactions to incidents.
• Offers a self-learning system that can continuously improve with evolving grid data and conditions.
• Increases transparency by providing intermediate outputs for each protection function, enhancing operator trust and system auditability.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The invention relates to a protection device (20) for monitoring a power supply grid (30, 70, 80), having a measured value detection device (21) for detecting measured values which specify the electrical state of the power supply grid (30, 70, 80), and an analysis device (22) which is connected to the measured value detection device (21) and is designed to carry out a plurality of protection functions and to make a decision as a result as to whether the power supply grid (30, 70, 80) is in an admissible or an inadmissible operating state. In order to be able to make a decision as to whether there is an admissible or an inadmissible operating state even faster and under different conditions of the power supply grid, according to the invention the analysis device (22) is completely designed in the form of a unique neural network (40) and has an input neuron layer (42) to which the measured values and/or values derived therefrom are supplied, at least one intermediate neuron layer (43a-n) and an output neuron layer (45) which outputs a classification of operating states of the power supply grid, wherein the neural network (40) is trained to carry out all protection functions together in order to classify the operating states. The invention also relates to a method carried out by means of such a protection device in order to monitor a power supply grid.