Deep Neural Network In Heat Demand Analysis

Heating load forecasting is great significant for the movement management of the heating system,environment and energy saving. It is one of the keys in development,operation and management of heating system, it is also a very important prerequisite for carrying on optimal adjustment for central heating system, so it has a direct contribution for the central heating system efficiency, energy saving and environmental protection.This paper makes an analysis of the status of central heating system, thorough analysis and studies the various factors that affect the heating load,it was proposed to the processing approach for influencing factors using quantitative of fuzzy data. On this basis, this paper used a new type of fuzzy neural network forecasting system. In the system, the sophisticated feedback BP network is used as the design core, and the quantitative parameters of fuzzy data of influencing factors are used as input values, and apply the neural networks to adjust and calculate the weights, then get the forecast network model through training and learning.After the model operation parameters determined, using MATLAB simulation to predict heat load, it shows that the forecast results of the fuzzy neural network is with higher accuracy and reliability, compared with that of BP network.