Factory Energy Monitoring Using ESI Distribution Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Factories face challenges in optimizing energy consumption due to difficulties in monitoring and analyzing abnormal energy consumption states, which can lead to energy waste and inefficiencies, particularly in determining the timing and factors contributing to these abnormalities.
Innovation Solution
An electronic device and method that utilize an energy stability indicator (ESI) and normalized energy stability indicator distribution (ESID) to compare energy consumption patterns, calculating distribution differences using relative entropy, and outputting observed variables and correlation coefficients to identify abnormal states and their causes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional energy consumption monitoring methods are used, then energy consumption data can be collected, but it is difficult to determine the timing and factors of abnormal energy consumption
Solution Approach 1:
The patent transforms energy consumption monitoring from raw data collection to statistical distribution analysis. By calculating energy stability indicators (ESI) and their distributions, the system changes the parameter representation from absolute energy values to normalized statistical parameters (mean, standard deviation, skewness, kurtosis), enabling effective detection of abnormal states and their timing.
Solution Approach 2:
The patent introduces energy stability indicators (ESI) as an intermediary between raw energy consumption data and abnormal state detection. The ESI and its distribution characteristics serve as a mediator that translates complex energy consumption patterns into analyzable statistical parameters, facilitating the identification of abnormal timing and contributing factors.
2Reliability
If detailed energy consumption data is collected for analysis, then abnormal states can be identified, but the complexity of the monitoring system increases
Solution Approach 1:
The patent extracts only the essential statistical characteristics (mean, standard deviation, skewness, kurtosis) from the complete energy consumption data distribution. This extraction approach maintains detection reliability by capturing the most informative aspects of energy consumption patterns while discarding redundant information, thereby simplifying the monitoring system.
Solution Approach 2:
The patent transforms complex energy consumption data into a standardized set of statistical parameters (four moments of the distribution). This parameter transformation reduces data dimensionality and complexity while preserving the essential information needed for reliable abnormal state detection through distribution comparison.
Data Source
AI summary
Disclosed are an electronic device for monitoring an abnormal state of energy consumption and a method of operating the same. The method includes receiving factory data related to product production of a factory and energy used in the factory, determining an energy stability indicator (ESI) indicating an amount of the energy used for the product production based on the factory data, generating a normalized energy stability indicator distribution (ESID) with respect to a preset time period based on the ESI, determining a distribution difference by comparing the generated ESID and a reference distribution predetermined for energy consumption of the factory, and outputting one or more observed variables for monitoring an abnormal state of consumption of the energy with respect to a period in which the distribution difference is large.