|
| |
|
|
Design of a system for detecting devices connected to the electrical network
Homola, Michal ; Kováč, Daniel (referee) ; Musil, Petr (advisor)
This master's thesis deals with the design of a system for detecting devices connected to power line network using the measurement of high-frequency noise through BPL (Broadband over Power Line) modems. The theoretical part involved familiarization with Power Line Communication (PLC), electromagnetic compatibility (EMC), impedance issues in PLC, and characteristics of noise in PLC. In the practical part, the suitability of the chosen PLC modems for the actual measurement was verified, followed by the measurement of temporal and spatial variability of network noise characteristics using these modems.For temporal variability, an experiment involving long-term measurement of refrigerator activity was conducted. For spatial variability, measurements were taken at multiple locations, with some locations serving as a training set and the remaining ones as a testing set. After selecting an appropriate machine learning model, the input data were feature engineered accordingly, followed by their evaluation.
|
|
|
Design of a system for detecting devices connected to the electrical network
Homola, Michal ; Kováč, Daniel (referee) ; Musil, Petr (advisor)
This master thesis deals with the creation of a system for detecting devices connected to the power network using the measurement of high-frequency noise obtained via BPL modems. In the theoretical part, there was an introduction to the issue of PLC, electromagnetic compatibility of EMC, the issue of impedance in PLC and noise characteristics in PLC. In the practical part, measurement of noise characteristics for individual devices and the creation of a dataset took place. The dataset, which was then tested on five machine learning models selected for this task based on their properties. Finally, the suitability of each model for our application was evaluated.
|
|
| |
guest ::
RSS feed.