Electric Grid Waveform Capture for AI Data Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electric power data collection systems are inadequate for gathering large datasets needed for machine learning and artificial intelligence applications due to limited data capture and connectivity issues, leading to insufficient raw data for advanced analytics and real-time situational awareness.
Innovation Solution
A system comprising transducers, data collection units, and a data concentrator that collects and processes electrical parameters like voltage, current, and power, with optional cloud-based analysis, enabling efficient data storage and transmission for machine learning and artificial intelligence systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional utility equipment is used to gather data, then aggregated measurements such as RMS voltage, current, and real power can be obtained, but raw waveform information is hidden and deeper analytics cannot be performed
Solution Approach 1:
The patent extracts raw waveform data from the aggregation process by implementing data collection units that capture individual waveform samples at multiple points in time, separating the raw data layer from the aggregated metrics layer to enable both types of analysis
Solution Approach 2:
The system transitions from single-point aggregated measurements to multi-dimensional waveform data by capturing voltage, current, and power at multiple time points within each cycle, adding temporal dimensionality that enables detailed waveform analysis
2Measurement precision
If advanced equipment captures raw waveform information, then deeper analytics become possible, but memory is limited and data capture is restricted to short bursts based on trigger conditions
Solution Approach 1:
The system performs preliminary data collection by continuously capturing waveform samples and storing them in circular buffers before any triggering event occurs, ensuring that data is ready for analysis regardless of when a disturbance occurs
Solution Approach 2:
The data collection units operate continuously rather than intermittently, maintaining persistent capture of waveform data across multiple cycles without relying on trigger conditions to initiate data collection
3Productivity
If continuous streaming of waveform data is implemented, then real-time analysis is enabled, but high speed always-on network connections are required which are unavailable at many points on the electric power grid
Solution Approach 1:
Instead of continuous streaming, the system implements periodic data transmission where waveform data is transmitted at scheduled intervals, reducing network bandwidth requirements while maintaining effective data collection and analysis capabilities
Solution Approach 2:
The system introduces intermediate processing and buffering at local devices before data transmission, allowing data to be collected and prepared locally without requiring constant high-speed network connectivity for the entire data collection process
4Quantity of substance
If large quantities of raw data are gathered from many points in the electric power system, then machine learning and AI algorithms can be trained, but the complexity of data collection and transmission increases significantly
Solution Approach 1:
The system segments the data collection function into distributed units at various points in the power system, with each unit responsible for collecting data locally, reducing the complexity of centralized collection while enabling comprehensive data gathering across the entire system
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Facilitates the collection and analysis of large datasets for machine learning and artificial intelligence systems, providing real-time situational awareness and predictive analytics for power systems, overcoming data capture and connectivity limitations.
Implementation Method 1
at least one transducer configured to measure an electrical parameter of a monitored element
Data Source
AI summary
A data acquisition and collection device, process, and/or system is disclosed that may be coupled with optional cloud-based data collection and analysis features. This device, process, and/or system provides for the collection of large datasets from the electric power grid to feed into ML/AI training systems, and optionally provides a mechanism to apply new algorithms back to the device, process, and/or system to enhance data collection and triggering mechanisms.