Smart Home Energy Management: VAE-GAN synthetic dataset generator and Q-learning

Mina Razghandi,Hao Zhou,Melike Erol-Kantarci,Damla Turgut

2023-05-15

Abstract:Recent years have noticed an increasing interest among academia and industry towards analyzing the electrical consumption of residential buildings and employing smart home energy management systems (HEMS) to reduce household energy consumption and costs. HEMS has been developed to simulate the statistical and functional properties of actual smart grids. Access to publicly available datasets is a major challenge in this type of research. The potential of artificial HEMS applications will be further enhanced with the development of time series that represent different operating conditions of the synthetic systems. In this paper, we propose a novel variational auto-encoder-generative adversarial network (VAE-GAN) technique for generating time-series data on energy consumption in smart homes. We also explore how the generative model performs when combined with a Q-learning-based HEMS. We tested the online performance of Q-learning-based HEMS with real-world smart home data. To test the generated dataset, we measure the Kullback-Leibler (KL) divergence, maximum mean discrepancy (MMD), and the Wasserstein distance between the probability distributions of the real and synthetic data. Our experiments show that VAE-GAN-generated synthetic data closely matches the real data distribution. Finally, we show that the generated data allows for the training of a higher-performance Q-learning-based HEMS compared to datasets generated with baseline approaches.

Machine Learning,Systems and Control

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to effectively generate synthetic time - series data to support energy consumption analysis and management in the case of the lack of high - quality public datasets for intelligent home energy management systems (HEMS). Specifically, the paper proposes a technique based on variational auto - encoder - generative adversarial network (VAE - GAN) to generate energy consumption time - series data for intelligent homes and explores the performance when this generative model is used in combination with Q - learning - based HEMS. Through this method, researchers hope to overcome the privacy and security challenges in actual data collection and at the same time improve the performance of HEMS, enabling it to operate more effectively in the actual environment. The key contributions of the paper include: 1. **VAE - GAN Generation Scheme**: Proposed a VAE - GAN - based scheme for generating synthetic time - series data of intelligent home energy consumption with high - time resolution. 2. **Performance Comparison**: Compared the performance of the proposed VAE - GAN method with methods based on Gaussian mixture model (GMM) and ordinary GAN. 3. **HEMS Performance Evaluation**: Trained a Q - learning - based HEMS model using the generated synthetic data and verified its performance in the real - world data environment through online testing. The results show that the data generated by the VAE - GAN method can enable HEMS to obtain higher profits than other baseline methods. These contributions not only solve the problem of data acquisition but also provide a new approach for the optimization of intelligent home energy management systems.

Smart Home Energy Management: VAE-GAN synthetic dataset generator and Q-learning

Variational Autoencoder Generative Adversarial Network for Synthetic Data Generation in Smart Home

Generating Synthetic Mixed-Type Longitudinal Electronic Health Records for Artificial Intelligent Applications

Synthetic Active Distribution System Generation Via Unbalanced Graph Generative Adversarial Network.

Synthetic Data Generation for Residential Load Patterns via Recurrent GAN and Ensemble Method

Smart Home Energy Management: Sequence-to-Sequence Load Forecasting and Q-Learning

Generative Adversarial Network-Assisted Framework for Power Management

Creating synthetic energy meter data using conditional diffusion and building metadata

Capturing multiscale temporal dynamics in synthetic residential load profiles through Generative Adversarial Networks (GANs)

Real-Time Energy Management in Smart Homes Through Deep Reinforcement Learning

A novel deep generative modeling-based data augmentation strategy for improving short-term building energy predictions

Internet of things dataset for home renewable energy management

Faraday: Synthetic Smart Meter Generator for the smart grid

Synthetic Generation of Patient Service Utilization Data: A Scalability Study

A synthetic dataset of Danish residential electricity prosumers

GAN-based Model for Residential Load Generation Considering Typical Consumption Patterns

Predicting short-term energy usage in a smart home using hybrid deep learning models

Optimization strategy based on deep reinforcement learning for home energy management

Electricity GANs: Generative Adversarial Networks for Electricity Price Scenario Generation

Home Electricity Data Generator (HEDGE): An open-access tool for the generation of electric vehicle, residential demand, and PV generation profiles

A Novel Two-stage Reinforcement Learning Framework for Sustainable Building Energy Management Systems