G m l Sistemler ile YGYI Kullanarak Ev Aletlerinin Enerji T ketiminin Takibi ve Enerji Tasarrufu nerilerinin Sunulmasi

This study introduces a semi-intrusive load monitoring system developed using embedded hardware to track the energy consumption of household appliances. The system collects data synchronously from five independent channels, allowing for accurate computation of power parameters. A two-stage filtering technique is employed to reduce transient effects and measurement noise. Devices with constant load profiles are identified directly, while appliances with more complex behaviors are recognized through inferential analysis. Unlike approaches that rely on machine learning, the proposed method offers a practical and efficient way to detect device activities on embedded systems with limited computational resources. Furthermore, it presents a novel strategy that reduces sensor count and system cost, while maintaining the ability to monitor appliances across multiple power lines without losing group-level associations.