Study On Energy Consumption Forecasting And Its Application To Energy Management

Energy is important substance foundation of humanity society, and every country or area's economy development need energy's support. Scientific and logical energy developme -nt stratagem can sustain local economy development, and it is based upon right investigating and judging energy situation. Energy consumption forecasting is an important part of energy planning. We can constitute optimum energy stratagem by energy consumption forecasting. Many researchers have forecasted energy consumption recent years, and different forecasting methods leaded to different results. Therefore, it is a target to establish an applicable and accurate forecasting method. This paper is composed of the below contents:Firstly, sustainable development system of energy was discussed. Grey model combined with partial least squares regression(PLS) was applied in forecasting the medium and long term city energy consumption for the first time. Single models and combined model were established and validated by using Dalian past years' data. The combined forecasting method possessed the two single method's advantages, and the result showed that forecast value was more accurate.Secondly, several independent variables were selected from economy, population, industry structure, energy structure and resident living aspects, and forecasted the primary energy consumption and final energy consumption in Dalian using the combined forecasting model. All these offered the reference for energy planning and decision-making.Finally, energy problems of Dalian were analyzed in energy safety, energy structure, energy efficiency and energy environment. The result showed that coal demand would be provided safely, and natural gas would be provided low safely because of single gas source in 2020 in Dalian. Energy structure of Dalian would have a great improvement, and energy efficiency would achieve a higher level in 2020. Emission quantity of SO₂, NO_x and soot would decrease after adjusting energy structure and industry structure, but there was a disparity between forecasting value and planning value. Strategic thought and advice was put forward basing forecasting and analyzing result.The model established is practical and universal, and can forecast energy consumption for specific area to select its influence factors. PLS overcomes mulriple relativity among independent variables, and explains dependent variable primely; GM(1,1) overcomes non -linearity of parameter. The combined forecasting method is more accurate to forecast...