Explainable Multi-Agent Recommendation System for Energy-Efficient Decision Support in Smart Homes

TL;DR

This paper presents an explainable multi-agent recommendation system for load shifting in smart homes, enhancing transparency and trust to promote energy-efficient behaviors among consumers.

Contribution

It introduces an explainability agent and improved predictive models incorporating weather data, advancing load shifting recommendations for household appliances.

Findings

Enhanced predictive accuracy with weather data integration

Improved transparency through explainability agent

Potential for scalable energy-saving recommendations

Abstract

Understandable and persuasive recommendations support the electricity consumers' behavioral change to tackle the energy efficiency problem. Generating load shifting recommendations for household appliances as explainable increases the transparency and trustworthiness of the system. This paper proposes an explainable multi-agent recommendation system for load shifting for household appliances. First, we provide agents with enhanced predictive capacity by including weather data, applying state-of-the-art models, and tuning the hyperparameters. Second, we suggest an Explainability Agent providing transparent recommendations. We also provide an overview of the predictive and explainability performance. Third, we discuss the impact and scaling potential of the suggested approach.