Intelligent soft-computing based modelling of naturally ventilated buildings

摘要: The paper presents recent results on the application of soft computing methodology for modelling internal climate in office buildings. More specifically, a part recently completed naturally ventilated building is considered which comprises three neighbouring offices and one corridor within Portland Building at University Portsmouth. approach adopted uses fuzzy logic modelling, neural networks adaptation genetic algorithms optimisation model. models are Takagi-Sugeno type built by subtractive clustering. As result latter, initial values antecedent non-linear membership functions consequent linear algebraic equations parameters determined. A method extensive search model structures presented fully explores dynamics plant. further adjusted back-propagation training network real-valued algorithm order to obtai...