Self-Organizing Map Load Feature Database for Electric Load Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying miscellaneous electric loads (MELs) in commercial buildings are inaccurate, robustness is lacking, and fail to distinguish similar loads, such as DVD players and set-top boxes, due to limitations in voltage and current characteristics measurement and processing.
Innovation Solution
A self-organizing map (SOM) load feature database is employed, using a plurality of neurons with weight vectors to identify electric load types by relating load feature vectors to the SOM database, based on sensed voltage and current signals, allowing for accurate classification and management of MELs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional voltage-current measurement and hierarchical clustering methods are used to identify MELs, then the system structure is simple, but the identification accuracy is insufficient and similar loads cannot be distinguished
Solution Approach 1:
The patent transforms the identification approach by changing from traditional voltage-current measurement parameters to a self-organizing map neural network model. This parameter transformation enables the system to capture complex nonlinear relationships in load signatures, significantly improving identification accuracy for similar loads while maintaining computational efficiency through the SOM algorithm's inherent dimensionality reduction capabilities.
Solution Approach 2:
The patent replaces traditional mechanical clustering methods (hierarchical clustering based on distance thresholds) with a neural network-based self-organizing map system. This substitution allows the system to automatically learn and adapt to load patterns without requiring manual threshold setting, thereby improving accuracy while the modular SOM structure keeps implementation complexity manageable.
2Adaptability or versatility
If load identification systems are implemented to manage MELs, then energy management capability is improved, but the difficulty of detecting and measuring load characteristics increases
Solution Approach 1:
The self-organizing map load identification system performs self-training and self-configuration by automatically learning load signatures from measured data. The system independently builds its classification model without requiring manual intervention for parameter tuning or threshold setting, thereby enabling versatile energy management while simplifying the detection process through automated adaptation to various load types.
Solution Approach 2:
The patent implements a preliminary training phase where the SOM system learns and stores load signatures before actual identification operations. This preliminary action prepares the system in advance, creating a ready-to-use classification model that simplifies subsequent real-time detection and measurement tasks, enabling efficient energy management without repeated complex analysis.
Data Source
AI summary
A method identifies electric load types of a plurality of different electric loads. The method includes providing a self-organizing map load feature database of a plurality of different electric load types and a plurality of neurons, each of the load types corresponding to a number of the neurons; employing a weight vector for each of the neurons; sensing a voltage signal and a current signal for each of the loads; determining a load feature vector including at least four different load features from the sensed voltage signal and the sensed current signal for a corresponding one of the loads; and identifying by a processor one of the load types by relating the load feature vector to the neurons of the database by identifying the weight vector of one of the neurons corresponding to the one of the load types that is a minimal distance to the load feature vector.