Modeling And Applications Of Grey Prediction Models Based On Energy Structures

Reasonable prediction of energy consumption is conducive to scientific control of energy consumption,adjustment of energy structure,and improvement of energy conservation and emission reduction monitoring and early warning system.But with the rapid development of the world,the energy mix is also changing.Therefore,some discrete data have relatively little reference value for energy consumption or future energy development trend,while the grey prediction model suitable for small sample modeling is just suitable for such problems.Based on the hierarchical structure,properties and consumption trend of unsaturated S-type energy system,three kinds of multi-variable grey prediction models based on energy structure are proposed.The effectiveness of the three models is verified by the actual regional energy consumption data in China.The main contents are as follows:First of all,the existing models pay little attention to the trend of variables themselves.This paper starts from the trend of energy consumption,based on the Logistic equation suitable for unsaturated S-type data,and combines the grey correlation analysis method to screen the influencing factors inside the energy system.Based on the difference information of differential and difference of grey theory,a multi-variable grey forecasting model of energy consumption based on Logistic equation is established.Secondly,since a single equation can not fully describe the complex internal relations of the energy system,based on the multi-variable gray prediction model of energy consumption of Logistic equation,this paper divides the whole energy system according to the energy properties and attributes,and establishes differential equations to reflect the complex relations of the energy system.The multi-variable grey prediction model was established from the system level,and the model was solved and applied,and the validity of the model was verified by case analysis.Thirdly,considering that the description of variable relationship may be affected if the coefficient of the grey prediction model is fixed,this paper optimized the method of solving the coefficient of the multi-variable grey prediction model of energy consumption based on the Logistic equation.The local polynomial regression of nonlinear optimization method was used to approximate the dynamic coefficient,and the weighted least square method was used to solve the dynamic coefficient,and the multi-variable grey prediction model of dynamic parameters was established.This makes up for the lack of rationality of the relationship between variables described by the multi-variable grey prediction model of energy structure and increases the robustness of the model.Finally,three kinds of multi-variable grey prediction models based on the development trend of energy consumption are applied to the prediction of fossil energy and clean energy consumption in actual regions of China,and analyzed with emphasis.The experimental results can provide reliable theoretical basis for China’s energy structure adjustment and energy security prevention.