Optimization Method Of Grey Prediction Model And Its Application Of China's Shale Gas Output Forecast

The grey prediction model is an important part of the grey system theory.However,the existing grey prediction models are still difficult to fully adapt to the development and changes of things,and the grey prediction models need to be continuously adjusted to meet actual needs.The commonly used single optimization method improves the modeling accuracy of the grey prediction model to a certain extent,but limits the ability of the model to improve its overall performance.How to improve the performance of the grey prediction model as a whole has very important research significance.China's shale gas output plays an important role in the formulation of energy policies.However,due to the very limited historical data of China's shale gas output,which shows a typical "small sample" characteristic,conventional machine learning based on large data sets methods and statistical methods cannot model small data sets.The grey prediction model is a method for modeling small sample data,but the traditional grey prediction model is inaccurate in predicting China's shale gas output.Therefore,the establishment of an effective grey prediction model has both theoretical and application values.At present,the grey prediction model optimization method is relatively simple,which limits the improvement of the overall performance of the model to a certain extent.Based on the idea of combinatorial optimization,this paper studies the optimization of the grey prediction model.The article conducts research work mainly from the following two aspects:(1)Research on optimization of grey prediction model.Taking the basic form of a typical threeparameter grey prediction model as the basis for model research,by simplifying its expression form,a general form of the three-parameter grey prediction model is constructed.Fractional accumulation / subtraction technology is introduced into the general form of the three-parameter grey prediction model,the integer sequence generation process of the modeling sequence is extended to fractional generation,and a fractional three-parameter discrete grey prediction model is established.Then based on the new information first principle,the time response of the model is derived.In the modeling process,particle swarm optimization is used to solve the cumulative generation order of the model based on the principle of minimum simulation error.Finally,the properties of the new model are discussed and proved.(2)Carry out research on the prediction of shale gas output in China.In the data pre-processing process,a qualitative analysis of future development trends and turning points was combined with a quantitative analysis of the grey shale gas output prediction model,and then the new model was used to model China's shale gas output from 2012 to 2018,and compared with the existing grey prediction model of shale gas.The results show that the new grey prediction method constructed in this paper is superior to the existing grey prediction method of shale gas output,and can achieve a better prediction of Chinese shale gas output.