Personalized Home BESS Recommender System Based on Neural Collaborative Filtering

Abstract

Battery energy storage systems have been becoming increasingly prevalent in residential sector. With more and more home battery energy storage system (HBESS) products appeared in the market, it would be difficult for the users to choose the most suitable battery product from the market. This paper proposes a personalized HBESS recommender system that provides decision-making support to residential users by recommending suitable HBESSs to them. The system integrates a neural collaborative filtering technique, which uses a General Matrix Factorization (GMF) model and a Multi-Layer Perceptron (MLP) neural network to infer the target user’s preferences on different HBESS products through analyzing the preference trends of a group of the target user’s similar users on a set of HBESS products. Based on this, the system generated a recommendation list of HBESS products to the target user. Numerical simulations based on real-world data are conducted to validate the effectiveness of the proposed system.