The Study on Building Energy Conservation Technology Based on Four-Layer Neural Network

Abstract

This paper build general regression neural network building conservation technology project appraisal model, Written software programs with MATLAB7.0 neural network toolbox to appraise building conservation technology project, trained and tested network by 17 ventilation samples, shows that GRNN have a good feasibility for appraisal. The main advantage of the GRNN model reflects four aspects. First of all, GRNN is a very simple and fast learning procedure thus it has less training time. Second, it is unnecessary to define the number of hidden layers or the number of neurons per layer in advance. Third, GRNN can handle linear and nonlinear data. Fourth, adding new samples to the training set does not require re-calibrating the model. Finally, it has only one adjustable parameter thereby making overtraining less likely. Because of these advantages, GRNN can be applied to many fields with other forecasting approaches or appraisal. GRNN building conservation technology project model building and training, testing can use neural network toolbox of MATLAB software.