Analysis Of Influencing Factors And Short-Term Load Forecast Of Gas Consumption In Urban Pipeline Based On Data Mining

At present,under the dual influence of Chinese economic ascent and environmental protection pressure,the domestic energy structure has been adjusted,which has greatly promoted the implementation of coal-to-gas conversion measures in residential,industrial and commercial fields.Urban gas demand has increased substantially,and the situation of supply and demand is more severe,"Gas shortage" and "guaranteed supply" have gradually become the keywords of urban gas enterprises.In the case of insufficient supply of upstream gas source indicators,how to plan the future development direction of urban gas market and formulate economic and reliable gas supply schemes has become a difficult problem faced by urban gas enterprises in order to ensure the smooth supply of gas to residents and industrial parks.Gas enterprises need to analyze the influencing factors of gas consumption in their cities to grasp the characteristics of gas consumption of urban gas users and the uneven characteristics of gas consumption,so as to optimize and adjust the gas consumption structure.On the other hand,it is necessary to study the short-term load forecasting and peak shaving schemes for urban gas in order to properly manage and purchase emergency gas sources and improve the construction of gas storage facilities.The influencing factors of urban gas consumption and the characteristics of short-term load forecasting are complex and changeable,and there are stochastic and uncertainties.It is necessary to start with its essential law and its own characteristics.Considering the limitation,incompleteness and complexity of the historical data of urban pipeline gas,this thesis uses data mining method to process and analyze the data.Taking a city gas enterprise in the northern region as an example,this thesis collects and collates the historical data of urban gas pipeline network in recent years,and uses correlation analysis to determine the influencing factors of urban gas consumption.Through processing and processing the relevant historical data,the potential relationship between the influencing factors of various types of user gas consumption in urban gas is excavated,which will help to clarify the mechanism of the formation of gas load and provide a basis for the prediction of urban gas load.This thesis focuses on practical application to help gas enterprises solve immediate problems.Firstly,it investigates a large number of basic data of gas consumption of H City Gas Company in North China,identifies the related factors of gas consumption,classifies these factors,and eliminates the factors that have very little impact on short-term load,and then uses the principal component analysis method to reduce the dimension of the main influencing factors data after pretreatment and remove the correlation between the data.The reconstructed data retains the main information of the original data.On this basis,data mining technology is used to analyze the main influencing factors and find out the law of the city's gas load.In view of the uncertain and non-linear characteristics of gas load system,this thesis chooses BP(Back Propagation)artificial neural network method to establish short-term load forecasting model.The extracted principal components are used as input variables of BP neural network,and the short-term load forecasting and verification are carried out with H city gas load data.Accurate prediction of short-term load of urban gas can help H city gas enterprises formulate reasonable gas supply plan,and it is also the basis of peak shaving and dispatching of gas enterprises.In addition,it also plays an important guiding role in urban gas planning and design.The innovations in this thesis are as follows: First,PM2.5(fine particulate matter)in air quality is taken into account when analyzing the influencing factors of urban gas consumption,There must be repeatability and high correlation between factors affecting PM2.5 and factors affecting gas load,so it is necessary to incorporate it into principal component analysis;Secondly,the combination of principal component analysis and BP neural network is used to predict gas consumption,which reduces the amount of calculation and improves the accuracy of prediction;...