Paper Conference

Proceedings of uSim Conference 2020: 2nd uSim Conference of IBPSA-Scotland


Estimating macroscopic occupant behavior with an end-use simulation model and smart meter data

Yusuke Kurokawa, Yoshiyuki Shimoda, Aiko Ikeda

Abstract: This study focuses on city-scale human behavior, using smart meter data and a bottom-up energy end-use simulation model to analyze macro-scale behavior such as going out on holidays and changing behavior due to weather conditions. With national time-use survey data as input, the end-use energy model simulates household occupant time-use schedules and energy-use behavior using attributes such as age and occupation, and estimates energy consumption for an appliance operation schedule based on these factors. By comparing the model’s estimated energy consumption with the actual smart meter data under specific conditions, such as bad weather and holidays, discrepancies likely attributable to differences between the responses given by those answering the survey and their actual behavior can be identified. For example, it was found that the level of electricity consumption indicated in the smart meter data was larger than the consumption estimated by the simulation model, both at midnight and on rainy days. This suggests that the bedtime of household residents is later than is indicated in their responses to the survey, and that the out-of-the-house rate decreases on rainy days. However, there are exceptions on special event days, such as Christmas Eve. Utilizing the bottom-up model makes it possible to examine city-level human behavior from a macro perspective using smart meter data that only provides energy use information. Understanding the characteristics of human behavior in various regions is important for efficient city-level energy management. By estimating human behavior at the macro level and revising the model using the results it produces, further improvements in the accuracy of the proposed city-level simulation model can be expected.
Pages: 142 - 147