Research On Building Energy Requirement Based On Multi-agent Model

Energy and environmental issues are important factors influencing social sustainable development.As the living environment of buildings has becoming the new foothold for pursuing life happiness,building energy efficiency should be complemented based on the satisfaction of occupants’ requirements.Discussing building energy efficiency from the perspective of occupant behavior is a new approach to discover the energy saving potential from the mechanism of building energy demand.Based on the occupant behavior of energy useage,a multi-level recoursive model of energy demand was established.The study object was the direct electricity demand of occupants in buildings.Following the orientation of energy demand generating and moving,the electricity demand can be described on the levels of single occupant,single building and multiple buildings.Because of the complexity of building coupling and energy systems themselves,the multi-agent system was adopted to model the electricity demand of the three levels.In order to obtain the time series of electricity demand,the occupant movement was considered as the information strea which connected the models of the multiple levels.A new university campus was taken as case study.The occupant behavior was divided into action and movement.Typical patterns of occupant energy usage actions were categorized based on survey data.On the occupant level,occpants behavior were divided into three periods of just arriving,normal working and soon leaving.The probability models was established to quantifiy the energy power of appliances which the actions made on.Quadratic polynomial was adopted to describe the typical bimodal or trimodal distribution of indoor occupant number of buildings.The stochastical fluctuation characteristics were modeled by Markov-Monte Carlo method.Finally,the model accuracy was further improved by Bayesian model,which depicting the occupants movement amoun multiple buildings in the campus.The theory of growth limiting of the energy consumption with occupant behavior was proposed on the building level.The limiting value of indoor occupant was determined based on clustering analysis.The piecewise functions of energy demand of single buildings with indoor occupant number were built.The energy demand of multiple buildings was the coupling of time series demand of single buildings.The model accuracy was proved by the the building cooling and heating load simulation based on Design Builder.Results of occupant behavior and building electricity loads were inputted to the software.The average error rate of simulated results with measured cooling and heating supplement values was about 5%.The energy demand quota of typical occupant behavior and buildings were calculated as applications of the proposed model.The calculated quotas were lower than the recommended values in relative standards for buildings with high occupant control probability.It can be concluded that higher energy saving potential can be realized when building energy system was managed based on the occupant demand method.