Energy Use Classification for Real-Time Abnormal Consumption Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current energy consumption analysis methods focus on daily averages, which do not provide detailed information about intervals of energy consumption throughout the day, failing to detect abnormal usage patterns and inefficiencies in real-time.
Innovation Solution
A method and system that collect energy consumption data, identify data clusters, categorize them, and analyze them to classify periods as normal or abnormal, enabling real-time management and providing recommendations for improvement, using a processor with energy meters to dynamically model normal energy consumption patterns and alert users to inefficiencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If daily average energy consumption values are used for analysis, then the analysis is simple and aggregated, but detailed information about energy consumption intervals throughout the day is lost
Solution Approach 1:
The patent segments energy consumption data into multiple intervals throughout the day (e.g., peak hours, off-peak hours, night hours) rather than using a single daily average. This segmentation allows detailed analysis of energy consumption patterns at different times while maintaining manageable complexity through systematic categorization of each interval.
2Ease of manufacture
If minimum energy consumption value is used as base load, then calculation is simplified, but accurate representation of actual base load is compromised
Solution Approach 1:
The system performs self-learning by automatically analyzing historical energy consumption data to identify and adapt to the actual base load pattern specific to each site. Rather than using a simplified minimum value, the system autonomously determines the true base load through continuous monitoring and pattern recognition, improving accuracy without requiring manual intervention.
3Productivity
If aggregated energy consumption values (sum, mean, median) are used, then reporting is simplified, but detailed information about energy consumption patterns is lost
Solution Approach 1:
The patent adds the time dimension to energy consumption analysis by breaking down aggregated values into temporal intervals. Instead of reporting only daily totals or averages, the system provides multi-dimensional data showing consumption patterns across different hours, days, and periods, enabling both efficient aggregation for high-level reporting and detailed analysis when needed.
4Reliability
If real-time energy consumption monitoring is implemented, then abnormal usage patterns can be detected, but system complexity increases
Solution Approach 1:
The system implements continuous feedback loops where energy consumption data is monitored in real-time, compared against learned normal patterns, and automatically triggers alerts or adjustments when deviations are detected. This feedback mechanism enables reliable abnormal pattern detection while managing complexity through automated responses rather than requiring complex manual intervention systems.
Data Source
AI summary
A method for energy management include steps of collecting data during a period of time, identifying data clusters within the data, categorizing the data clusters in the period of time, analyzing the data clusters, and assigning a classification to the period of time. Real time data can be compared to scheduled performance. Real time notifications can be provided to a user where energy is being used inefficiently. Energy efficiency of meters within a site can be ranked to identify meters where energy is being used inefficiently. A system is also provided for performing the method of the present disclosure.