Device Identification via Predicted and Detected Power Consumption
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current appliance load monitoring techniques face challenges in accurately identifying and distinguishing between different versions of active devices within a premises using power consumption data, particularly in non-intrusive methods, which can lead to inefficiencies in device management and security updates.
Innovation Solution
A method involving a device identification system that uses a combination of predicted and detected identities based on power consumption data, processed using machine learning techniques such as LSTM neural networks and hierarchical support vector machines, to determine the specific type and version of active devices, and adjusts confidence scores for robust identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If non-intrusive load monitoring (NILM) techniques are used to identify devices, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The patent segments the device identification process into two distinct components: a predicted identity determined from historical power consumption data using machine learning, and a detected identity determined from current power consumption data. This segmentation allows the system to combine the advantages of both predictive modeling and real-time detection, improving overall identification accuracy while maintaining the ease of non-intrusive monitoring.
Solution Approach 2:
The system performs preliminary action by determining a predicted identity of the active device based on historical power consumption data before final identification. This prediction serves as a preliminary step that guides the subsequent detection process, allowing the system to prepare expected device characteristics and compare them against actual measurements, thereby improving measurement precision while maintaining ease of operation.
2Productivity
If appliance load monitoring techniques are used to detect device operation, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The patent segments the identification process into predicted identity and detected identity components, allowing the system to process power consumption data through multiple analytical pathways simultaneously. This segmentation enables the system to maintain high productivity by processing data efficiently while improving measurement precision through the combination of predictive modeling and detection algorithms.
Solution Approach 2:
The system performs preliminary analysis by determining predicted identity from historical data before final device identification. This preliminary action enables the system to pre-process and pre-analyze power consumption patterns, improving productivity by reducing the computational burden during real-time detection while enhancing measurement precision through comparative analysis.
3Measurement precision
If machine learning techniques are used to process power consumption data, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the machine learning processing into distinct functional components: one component determines predicted identity from historical data, another determines detected identity from current data, and a third component combines these results. This segmentation makes the complex machine learning system more manageable and easier to implement while maintaining high measurement precision through specialized processing in each component.
Data Source
AI summary
A device identification method, a device identification system and a device prediction component. The method can include determining, based on first power consumption data indicative of a first power consumption associated with a premises within a first time period, a predicted identity of an active device at the premises within a second time period subsequent to the first time period. A detected identity of the active device at the premises within the second time period is determined, based on second power consumption data indicative of a second power consumption associated with the premises within the second time period. A determined identity of the active device at the premises within the second time period is determined, based on at least one of the predicted identity and the detected identity.