Research On Demand Forecasting Model Of Coal Logistics Based On Drosophila-Grey Neural Network

As a developing country,China's economic development will depend on stable energy supply for a long time.The amount of coal stored is the highest among all kinds of energy in China,and the development speed of clean new energy is slow.This determines that the main energy consumption in China is still coal in the future.China's coal production and consumption areas are very asymmetrically distributed.Coal needs long-distance transportation to reach the final consumption area.Therefore,coal logistics system is an important bridge to ensure the stability of coal supply and demand.The forecast of coal logistics demand is the premise of optimizing and integrating coal logistics resources,which is conducive to improving the efficiency of coal logistics system.In this paper,based on the analysis of the current situation and trend of energy consumption in China,grasping the situation of coal consumption,consulting and studying the relevant research of coal logistics at home and abroad,the demand forecasting model of coal logistics is studied.Firstly,under the premise of studying books and consulting the literature,the theoretical knowledge of coal logistics,logistics demand and demand forecasting are compiled.On this basis,the index system of coal logistics demand forecasting model is constructed.Referring to the index system construction standards,considering the economic development,industrial structure,energy consumption structure,national consumption level,environmental protection policy and other factors,combining with the availability and uniformity of relevant data,and finally selecting the gross national product,output value of the secondary industry,total coal production,total energy consumption,total coal import and export,per capita consumption expenditure of residents,the number of urbanized population,and the proportion of cities with annual precipitation in China as the influencing factors.Considering the ultimate result of coal transportation is consumption,so coal consumption is selected as the measure index of demand forecasting.Then focus on the prediction model.Comparing the research status of related fields at home and abroad and select combination model.Based on the characteristics of small number of research samples,complex influencing factors and a certain trend in the development of things,this paper selected to combine grey model and BP neural network.And the emerging fruit fly algorithm optimization model is adopted is adopted to improve the connection weights and thresholds for the case that the neural network convergence speed is slow and falls into local optimum.Comparing the relative errors of GNNM,FOA-GNNM and GA-GNNM,it shows that FOA-GNNM has higher prediction accuracy and is suitable for short-term prediction of coal logistics demand.Finally,FOA-GNNM is used to predict the coal consumption in 2018-2022.Based on historical data and relevant policy planning,the proportion of coal transportation by railway,highway,sea and other modes is calculated,and then calculating the coal transportation of various modes of transportation in 2018-2022 by using consumption.Finally,according to the volume forecast results,some suggestions are put forward on the coal logistics volume arrangement,line planning and reserve system.Figure 7 Table 14 Reference 60...