Research And Application On Analysis Model Of Large Public Energy Consumption Based On Genetic Bp Neural Network Technology

Relying on the construction projects of Guangxi University economical campus building energy saving monitoring platform, this paper constructed the energy monitoring platform of Guangxi University saving campus, and researched the general framework, data transmission, design of the measuring point, sub-monitoring model, selection of equipment, hardware installation and commissioning, energy monitoring software platform function of the energy monitoring platform, and in the process of using energy saving monitoring platform, it was found that the quota of energy of energy settlement system lacking scientific basis, which cause a big waste of energy distribution of the school.To solve this problem, in this paper, BP neural networks can solve nonlinear problems, with the global approximation and generalization ability of the large-scale public building energy analysis model program. Through the analysis of the factors affecting changes in building energy consumption, this paper used the gray correlation analysis method and determined the main factors that affect building energy consumption of large public buildings in Guangxi University, including temperature, humidity, weather characteristics, whether days. These are the input of the model, the output is building energy consumption, thus, a analysis model of building energy consumption was established based on BP neural network. However, the BP neural network has slow convergence speed and is easy to fall into local minimum problems in practical application. In view of this, genetic algorithm was used to optimize the weights and thresholds of BP neural network. The improvement of large public building energy analysis model was completed, and visualization of large public buildings energy consumption of the analysis model of the interface is achieved. Test results of model show that the accuracy of analysis model of large-scale public buildings energy consumption based on genetic BP neural network in this paper is relatively high, compared with the traditional BP neural network model, the result is more accurate, the relative error of analysis model was reduced from10.08%to5.25%. It was proved that genetic BP neural network can be better applied to the research of analysis model of large scale public buildings energy consumption in Guangxi University, and be applied in large-scale public building energy consumption fixed and make foundation for the analysis of quota of energy...