Energy Consumption Alerting via Data Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional energy consumption measurement systems are inadequate for real-time monitoring and control, especially in large sites with multiple electrical appliances, as they lack the granularity to detect unusual energy consumption patterns efficiently, making it difficult to maintain energy efficiency and meet energy targets.
Innovation Solution
An energy consumption alerting method and system that uses sensors to measure location-specific energy consumption values, decomposes them based on seasonal and daily patterns, and compares these values with reference data to detect outliers, notifying users of unusual consumption through a cloud-based platform.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual metering devices are used for each electrical appliance, then measurement precision is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent segments energy consumption data by decomposing aggregate measurements into individual appliance contributions using machine learning models. Instead of physically segmenting the measurement system into multiple meters, the system virtually segments the data through algorithms that attribute consumption to specific appliances based on electrical signatures and usage patterns.
Solution Approach 2:
The patent introduces machine learning models and data processing algorithms as intermediaries between the aggregate metering device and the user. These intermediaries analyze the raw consumption data, decompose it into appliance-level insights, and present actionable information without requiring physical installation of multiple meters.
2Measurement precision
If individual metering devices are deployed throughout the site, then measurement precision is improved, but loss of time for installation and data collection increases
Solution Approach 1:
The patent creates virtual copies of individual appliance metering functionality through software algorithms. Instead of installing physical metering devices at each appliance, the system uses machine learning models that replicate the measurement and analysis capabilities of individual meters through software-based decomposition of aggregate consumption data.
Solution Approach 2:
The patent makes a single aggregate metering device perform multiple functions by combining it with machine learning algorithms. The system not only measures total consumption but also automatically decomposes it into individual appliance contributions, providing both aggregate and detailed measurement capabilities through a unified system.
3Device complexity
If only aggregate energy consumption data is collected, then device complexity is reduced, but loss of information about individual appliance consumption increases
Solution Approach 1:
The patent performs preliminary data processing and decomposition using machine learning models before presenting the information to users. The system pre-analyzes aggregate consumption data to extract appliance-level insights, trends, and anomalies, making detailed information available without requiring complex user-side analysis tools.
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously monitors appliance-level decomposed data and provides actionable insights to users. The machine learning models learn from consumption patterns and provide feedback about abnormal usage, energy-saving opportunities, and appliance performance, transforming raw data into meaningful information.
Data Source
AI summary
An energy consumption alerting method includes measuring location-specific energy consumption values over a specific period at a sensor deployed at a location of a monitored site, and decomposing the location-specific energy consumption values according to a first characteristic. The method further includes decomposing the location-specific energy consumption values according to a second characteristic, and obtaining a first decomposed energy consumption value. The method additionally includes determining a corresponding first reference value based on the decomposed values, and comparing the first decomposed energy consumption value with the determined corresponding first reference value. Additionally, the method includes notifying a user if the first decomposed energy consumption value and the determined corresponding first reference value differ from each other. Furthermore, an energy consumption alerting system and a cloud-based energy consumption alerting platform is provided.