Smart Meter Hot Socket Prediction via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Older utility meter sockets may deteriorate over time, leading to poor electrical connections and the potential for hot sockets, which can cause electrical arcs and even house fires. Existing methods for detecting hot sockets rely on temperature measurements that may not detect arcs in time to prevent damage.
Innovation Solution
A machine learning-based approach that predicts the occurrence of hot sockets by analyzing historical data from smart meters, including daily electricity consumption, meter events, service orders, and weather data, to generate features that are used to train a predictive model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If temperature measurement methods are used to detect hot sockets, then detection capability is achieved, but detection timing is too late to prevent damage
Solution Approach 1:
The system performs preliminary actions by collecting and analyzing multiple predictive features (consumption patterns, event history, service orders, weather data) before the hot socket condition fully develops. The machine learning model predicts potential hot socket occurrences in advance, enabling preventive maintenance before actual overheating or arcing occurs, thus resolving the timing delay inherent in traditional temperature-only detection methods.
2Loss of time
If machine learning prediction is implemented, then detection timing is improved, but system complexity increases
Solution Approach 1:
The system segments the prediction task into distinct feature categories (consumption data, event history, service orders, weather information) that are processed separately and then integrated by the machine learning model. This segmentation allows for modular data collection and processing, making the complex prediction system more manageable and maintainable while still achieving early detection capabilities.
Data Source
AI summary
Embodiments predict an occurrence of one or more hot sockets among a meter installation of a plurality of smart meters. Embodiments receive historical data over a predefined time period from the plurality of smart meters, the historical data including, for each smart meter, an amount of daily consumption of electricity for the meter, historical meter events for the meter, and all service orders associated with the meter. Embodiments pre-process the historical data to generate hot socket features. Embodiments train a machine algorithm using the hot socket features, and use the trained machine algorithm to predict the occurrence of one or more hot sockets.