HYDROSAFE: A Hybrid Deterministic-Probabilistic Model for Synthetic Appliance Profiles Generation

Area of Science:

• Energy Systems
• Data Science
• Artificial Intelligence

Background:

• Realistic appliance power consumption data are crucial for smart home energy management systems and algorithms.
• Existing publicly available datasets are limited and costly to acquire.
• A need exists for efficient methods to generate high-fidelity synthetic power consumption data.

Purpose of the Study:

• To propose HYDROSAFE, a novel hybrid deterministic-probabilistic model for generating synthetic appliance power consumption profiles.
• To address the scarcity and time-consuming nature of collecting real-world power consumption data.
• To enhance the development of smart home energy management systems through realistic synthetic data.

Main Methods:

• Employed the Median Difference Test (MDT) for characterizing power consumption profiles.
• Utilized Density and Dynamic Time Warping based Spatial Clustering for appliance operation modes (DDTWSC) to cluster appliance usage.
• Integrated stochastic elements (white noise, switch-on surge, ripples, edge position) for enhanced realism.

Main Results:

• The HYDROSAFE model successfully generated synthetic appliance power consumption profiles.
• Evaluation using a normalized DTW-distance matrix demonstrated high fidelity.
• Achieved an average DTW distance of ten samples at a 1Hz sampling frequency, indicating close mimicry of real-world data.

Conclusions:

• HYDROSAFE effectively generates realistic synthetic appliance power consumption data.
• The model's high fidelity supports its application in developing and testing smart home energy management systems.
• This approach offers a viable solution for overcoming data limitations in the field.