Machine Learning - Genetic Algorithm Optimization
Team : Elisabeth Kathryn, Anastasia Mimosa
Environmental simulation was integrated into the early stage of form finding. Genetic algorithm solver was used to test and find the most optimum parameter combination in order to get the best building performance. Several parameters tested in this experiment were the building’s four horizontal section profiles center point location, dimension and ratio. Whereas, the profile’s shape and height, also the total volume of the tower were all fixed. Solar radiation analysis was done for both facade and roof surfaces of the tower. The objective of the first experiment was to minimize the solar radiation which is applicable when the architect needs to reduce the building’s heat gain and reduce the cooling load. Then the objective of the second experiment was to maximize the solar radiation which, for instance, can be applied when the building envelope is built from photovoltaic panels and wants to receive as much energy as it can. The optimization results from two different shapes of profile (oval and rectangle) show similar tendencies.
Commenti