NodePM: a remote monitoring alert system for energy consumption using probabilistic techniques

Area of Science:

• Electrical Engineering
• Computer Science
• Data Science

Background:

• Smart grids require efficient monitoring of electronic equipment.
• Detecting novelties in energy consumption is crucial for grid management.
• Existing methods often lack real-world validation and efficiency.

Purpose of the Study:

• To propose an intelligent method, the Novelty Detection Power Meter (NodePM), for detecting novelties in smart grid-monitored electronic equipment.
• To evaluate the effectiveness of NodePM in reducing power consumption and improving novelty detection accuracy.
• To validate the method through real-world experiments.

Main Methods:

• Calculating device entropy using a Markov chain model.
• Employing a machine learning algorithm for novelty identification.
• Integrating NodePM into a wireless sensor network (WSN) for remote energy monitoring.

Main Results:

• NodePM reduced monitored equipment power consumption by 13.7%.
• NodePM demonstrated superior novelty detection efficiency compared to a literature-based approach.
• The method was validated in real-world experimental environments.

Conclusions:

• NodePM is an effective intelligent method for detecting novelties in smart grids.
• The proposed method offers significant power savings and improved detection accuracy.
• Real-world experimental validation confirms NodePM's practical applicability and performance.