Modeling and Simulation of Individual User Behavior for Building Performance Predictions

摘要

User behavior is a key to the correct prediction ofrnbuilding performance. Performance is meant here not only inrnrelation to energy consumption, but more importantly to userrnsatisfaction and, in commercial buildings, job performance.rnThe contributions of this paper are static and dynamic modelsrnof individual users and user activities in an office environmentrnin relation to building control systems. The models arernbased on a well structured general model of five different domainsrnof building entities, including user activities. Modelsrnfor specific simulation projects are refinements of this generalrnmodel. It is shown how multi-agent based simulations canrnbe automatically derived from such models. A case study ofrnan university office with irregular occupancy is introduced asrnan example.