Expert system for the calculation of building´s design cooling load based on artificial neural networks

In the stage of schematic design, some data needed to calculate design cooling load can not be supplied with accurate values, designers have to assess a value according to their experience only. The calculation of a building´s design cooling load is different from the load analysis of existing buildings, it is a prediction of the maximum possible load of a future building when it is put into use, so it has the characteristic of uncertainty and prediction. In order to improve calculation efficiency and enhance accuracy, expert system technology is introduced and incorporated with a load index assessment method. Consequently, an expert system for the calculation of a building´s design cooling load is established on the basis of artificial neural network theories. Application examples indicate that this system is simple and workable; the calculation results thereof achieve high accuracy, and calculation methods are in conformity to the thinking habits of designers.