Artificial neural network for predation of cooling load reduction using green roof over building in Sustainable City

Abstract Over the summer's two test structures one with green roofs and other with RCC was built and tested at Sustainable City, Ujjain to determine their cooling potential. Results indicate that the test cell with the green roof consistently performs better than the test cells with the conventional cement RCC roof. The objective of this work is to train an artificial neural network (ANN) to learn to predict the reduction in heat gain from the roof buildings with the different experimental data. A number of different training algorithms were used to create an ANN model. This study is helpful in finding the thermal comfort and energy saving of building by applying green roof over the roof. The data presented as input were daily Statistics for Dry Bulb temperatures temperature, relative humidity, average solar intensity and wind speed. The network output was reduction in heat gain from roof. The advantages of this approach compared to the conventional algorithmic methods are (i) the speed of calculation, (ii) the simplicity, (iii) adaptive learning from examples and thus gradually improve its performance, (iv) self-organization, (v) real time operation. ANN gives satisfactory results with deviation of 4.7% and successful prediction rate of 93.8–98.5%.