Research On Energy Data Quality Analysis Model Based On Regional Convergence

Data resources have become important production factors.As the basis of national economic and social,the accuracy of energy statistics continues to increase.China is the largest energy consumption country,energy consumption determines structural change of energy industry and energy data quality is related to the development of economy and society.However,the energy statistics system in China is not yet perfect,and the underlying statistics are not complete enough.Energy statistics problems will make research inaccurate and policy making deviated.Therefore,it is of great significance to analyze the quality of the energy statistics.This study constructs a set of universally applicable analysis model of energy statistical data quality.Selecting the panel data of energy consumption in 30 regions of China from 1997 to 2017,ARMA model and the joint estimation method are used to identify outliers in time series.China’s energy consumption level continues to rise,especially in heavy industrialization stage.Take partial model test results as example,the calculated results of energy consumption in Shandong province in 2004 and coal consumption in Hubei province in 2013 are 3.14 and 3.30 respectively,both of which are greater than the critical value and have sudden outliers.In order to analyze spatial geographic distribution characteristics,this paper combines regional convergence and spatial econometric model,adds spatial effects to construct spatial convergence model.The results show that unconditional β convergence rate of China’s energy consumption is 1.94%,and conditional β convergence rate is 6.25% by selecting five influence factors: economic level,industrial structure,fixed asset investment rate,degree of opening and rate of technological progress.Due to the uneven development between regions,this paper uses logt test method to classify convergence clubs,studies the internal development mechanism of the clubs.Energy consumption is divided into two clubs and coal consumption is three.After analyzing differences within and among clubs,the quality of energy consumption data is analyzed on the spatial fixed effect index,which can explain the influence of spatial geographical factors and indirectly judge the quality of energy consumption data.The spatial panel model is constructed of each region,the estimated value in the corresponding year is given through the model fitting result,which fluctuates is reduced compared with the actual consumption.Due to the special historical background and policy influence,statistical data may appear outliers.This study provides feasible suggestions for improving the quality of energy statistical data,which is conducive to carry out better energy research and forecasting work.