Load Feature Vector Distance Calculation for Electric Load Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying miscellaneous electric loads (MELs) in commercial buildings lack accuracy and robustness, often grouping similar loads together based on active and reactive power consumption, failing to distinguish between dynamic loads, and requiring steady-state operation, which hinders effective energy management and savings.
Innovation Solution
A method and system that utilize a load feature database to identify electric load types by determining the minimum distance of a multi-dimensional load feature vector to a database of known load types, employing sensors to sense voltage and current signals and a processor to calculate distances using features like true power factor, current harmonic distortion, and V-I trajectory characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional power metrics and eigenvectors are used for load classification, then the classification process is simplified, but the accuracy of distinguishing similar loads deteriorates
Solution Approach 1:
The patent transitions from traditional two-dimensional power metrics (active and reactive power) to a multi-dimensional feature space incorporating voltage-current trajectory, harmonic content, and transient characteristics. This dimensional expansion enables better separation of similar load types that occupy overlapping regions in traditional power planes, directly resolving the contradiction between simplicity and accuracy.
Solution Approach 2:
The patent changes the parameter set used for classification from conventional power metrics to a comprehensive set including voltage-current trajectory parameters, harmonic distortion indices, and transient response characteristics. This parameter transformation allows the system to capture subtle differences between similar loads while maintaining a systematic classification approach.
2Stability of the object's composition
If hierarchical clustering with fixed distance thresholds is used, then the classification structure is well-organized, but the ability to distinguish dynamic loads deteriorates
Solution Approach 1:
The patent introduces dynamic analysis by examining the time-varying characteristics of voltage and current trajectories, harmonic content evolution, and transient responses. This dynamic approach allows the classification system to adapt to changing load conditions while maintaining the hierarchical structure, resolving the contradiction between structural stability and adaptability to dynamic loads.
Solution Approach 2:
The patent performs preliminary classification based on steady-state characteristics to establish the hierarchical structure, then applies dynamic feature analysis to refine the classification of similar loads. This two-stage approach maintains organizational stability while enhancing the ability to distinguish dynamic and transient load behaviors.
3Difficulty of detecting and measuring
If steady-state operation requirements are imposed, then measurement and analysis are simplified, but the applicability to real-world dynamic loads deteriorates
Solution Approach 1:
The patent employs periodic sampling of voltage and current waveforms to capture both steady-state and transient characteristics. By analyzing periodic patterns in the time-domain and frequency-domain data, the system simplifies measurement requirements while maintaining applicability to dynamic loads, resolving the contradiction between measurement simplicity and real-world applicability.
Solution Approach 2:
The patent performs preliminary steady-state identification to establish baseline characteristics, then uses transient and dynamic feature extraction to handle non-steady-state conditions. This approach maintains the simplicity of steady-state analysis while extending applicability to dynamic real-world loads through additional dynamic feature layers.
Data Source
AI summary
A method identifies electric load types of a plurality of different electric loads. The method includes providing a load feature database of a plurality of different electric load types, each of the different electric load types including a first load feature vector having at least four different load features; sensing a voltage signal and a current signal for each of the different electric loads; determining a second load feature vector comprising at least four different load features from the sensed voltage signal and the sensed current signal for a corresponding one of the different electric loads; and identifying by a processor one of the different electric load types by determining a minimum distance of the second load feature vector to the first load feature vector of the different electric load types of the load feature database.