Short-term Forecasting Model,Optimization Algorithm And Application Of International Crude Oil Price

As a worldwide problem,international crude oil price forecast has been paid great atten-tion by scholars and oil companies at home and abroad.Such problems involve mathematics,energy economy,engineering,control theory,computer science,etc,and are an emerging inter-disciplinary subject.Based on the analysis and prediction of the international crude oil market,this paper summarizes the international crude oil market and its price formation mechanism,mainly studies the short-term prediction model of international crude oil price,optimizes the algorithm and its application.Applying the theoretical knowledge of the econometric model such as integer planning,multi-regression and the variational systems,the two-layer random integer planning model of international crude oil price forecast and the historical optimal time-matching model based on the influence factors of international crude oil price were established,and the discrete linear time lag non-homogenous dynamic system of domestic diesel wholesale sales was constructed.The strong stability of discrete linear time lag non-homogenous dynamic system is proved.On the one hand,this paper puts forward the short-term international crude oil price trend prediction model,realizes the combination of the theory and practice of model prediction,greatly enhances the interpretation of the short-term international crude oil price pre-diction results;On the other hand,by constructing the multi-regression model,the short-term international crude oil price forecast results are applied to the actual price marketing strategy of oil companies,which not only enhances the science of price marketing strategy,but also en-hances the practicality of the short-term international crude oil price forecast results.The main content of this paper is summarized as follows:(1)To summarize the international crude oil market and its price formation mechanism,including the main international crude oil futures market,the international crude oil price for-mation mechanism of the historical evolution process.Based on the study of the law of the trend of the short-term international crude oil price itself,the two-tier random integer planning mod-el of international crude oil price prediction with the optimal target of the international crude oil price prediction error expectation and variance is constructed,according to the optimal state transfer matrix obtained by the model,and predicts the future short-term international crude oil price trend according to the principle of maximum probability.Empirical analysis shows that the international crude oil price prediction error obtained by this method is relatively small,and has some practical value in predicting the short-term international crude oil price trend,domestic refined oil price adjustment and so on.(2)We select the representative US dollar index,the gold price and the Dow Jones index as an external market environment data indicator that simulates the changes in international crude oil price movements,and construct the best similar time period model algorithm in history.Using this optimization algorithm to obtain the best historical period of the data indicator with the change of the trend of the data indicator in the specified period,by comparing the historical best similar period of different data indicators,looking for the intersection or coincident part of the similar period,the study and analysis of the change of the international crude oil price trend in the comparison time period and the specified time period,and forecasting short-term international crude oil price movements.(3)According to the mechanism of the formation of domestic refined oil prices,based on the existing short-term international crude oil price forecast results,a discrete linear time lag non-homogenous power system of the wholesale sales volume of diesel oil enterprises has been constructed,and the optimal sales efficiency of petroleum enterprises is established as the objec-tive function,which proves that there is an optimal solution to the problem of optimization of the target function.The optimization algorithm of diesel wholesale price adjustment is construct-ed,the influence of different price marketing strategies on the wholesale sales and efficiency of diesel oil enterprises is analyzed quantitatively,and the best sales efficiency can be realized for petroleum enterprises to seize the opportunity of oil price fluctuations,and provides scientif-ic and reasonable price marketing strategy suggestions,which have good practical application value.(4)According to the linear variation system and the limitation of the basic matrix solution of the non-homogenous linear variation system,we prove the strong stability of the discrete linear time lag non-homogenous power system of diesel sales.