Analysis Of Energy Consumption Characteristics And Energy Consumption Prediction Method

The planning of the energy demand analysis is the most important basis for configuring and optimizing the energy system.There are two principles need to pay attention to: Firstly,it is necessary to subdivide and calculate the terminal energy consumption for various uses.Secondly,it is necessary to making dynamic modeling based on spatial distribution and time.However,the existing energy consumption simulation analysis methods have their own limitations,it can not accurately accomplish this goal.This paper carries out a series of analysis and researches for the historical energy consumption data of the existing regional buildings and constructs the index system of regional buildings energy consumption.Based on this,a model of building energy consumption forecasting based on BP neural network is proposed,and the regional energy system is configured and optimized.However,there are many factors influencing the regional buildings energy consumption,and the methods of establishing energy consumption models are various.How to choose the appropriate influencing factors and modeling methods is the focus and difficulty of this paper.This paper summarizes the research status of building energy consumption and energy consumption prediction model in the world,the paper also analyze the applicability of different energy consumption predicting methods to this subject.It is concluded that the bottom-up and BP neural network method are more suitable for establishing the regional energy consumption forecasting model in the planning stage.After the research method is established,the paper has carried out the following work:First of all,the first step is to use the expert investigation method to screen many factors that affect the building energy consumption,screening out nine factors.The second step adopts the analytic hierarchy process to analyze the weight of nine factors,and finally get the five indicators which are most relevant to the building energy consumption.Five indicators are as follows,the location,the building type,the construction area and the energy intensity.Take these as basis for the classification of the construction work.Secondly,this paper builds analysis and model for existing regional building historical energy consumption data with BP neural network algorithm.And the paper uses MATLAB to accomplish this process and tested of the model.In addition,this paper summarizes the typical cold and heat sources as well as their scope of application,and puts forward three different regional energy system configurations:conventional cold and heat source systems,distributed energy systems,and regional energy centers,and its scope of application was analyzed.Finally,the paper predicts the need of using energy demand based on real area of the cold areas.On this basis,the paper choose different number of energy center to configure regional energy system,and compare the initial investment,operating costs and operating characteristics of different programs.The point is to provide a reference for the allocation of regional energy systems.