[4I2-GS-2] Machine learning: Living with AI

[4I2-GS-2-02] Estimation Method of Household Electricity Consumption using AutoEncoding Deep Learning

〇Isao NAKAYAMA¹, Takeo YAGIHASHI¹, Yoshinobu KAWAHARA¹ (1. Tokyo Electric Power Company Holdings, Inc.)

Keywords:deep learning, Energy, auto encorder

Electricity consumption and usage patterns in ordinary households vary widely depending on the resident’s family composition and lifestyle, but the air conditioner heating is used when the outside temperature is low, and water heaters such as EcoCute automatically works at night. It can be considered that the driving patterns are combined by a driving pattern of a plurality of common devices to some extent. In this paper, in order to estimate the power consumption of each household for each application, the actual power consumption is machine-learned by an AutoEncoder, a common pattern is extracted, and the accuracy is verified. As an example, the result of estimating the power consumption of EcoCute is shown.