| In recent years,the development of Tibet’s freight transport industry has made certain achievements,with the volume of freight transport constantly expanding,freight transport infrastructure conditions improving significantly,and the freight transport circulation system becoming sounder.It has made positive contributions to the promotion of Tibet’s sustained and stable economic development.However,the growth of the freight industry in Tibet still faces the problems such as unreasonable transport structure,limited freight capacity and high logistics costs.These problems limit the enhancement of the level of freight transport services and are not beneficial to the steady growth of the logistical industry in Tibet.In a bid to enhance the development environment and solve the development problems of the freight industry in Tibet,it is necessary to grasp the development trend of the freight industry in the future and reasonably plan the transportation system.And freight demand forecast is the pioneering work to achieve these goals.This paper takes the demand for freight transport in Tibet as the research object.Research related to the analysis of freight demand influencing factors and the selection of freight demand forecasting methods based on the identification of freight demand characteristics.This paper firstly analyzed and reviewed the research of the freight demand forecasting at home and abroad in recent years,and then showed the significance of this paper.Then it elaborated on the relevant theories of transport demand and freight demand,and the essential connotation of freight demand was profoundly grasped to lay the theoretical foundation for freight demand forecasting.Later,it analyzed the freight demand of Tibet,including exploring the particularity of the freight demand of Tibet,analyzing the factors affecting the freight demand of Tibet from qualitative and quantitative perspectives.Finally,the main influencing factors of Tibet’s freight demand and the order of influence size were obtained.The top three influencing factors were: added value of the secondary industries,total retail sales of consumer goods and output of major industrial products.In the end,the freight demand forecasting model was built through the above analysis.A BP neural network model,a support vector machine model and their combined model were developed to forecast freight traffic in Tibet from 1991 to 2021 with the help of Python and MATLAB software.Then compared with other forecasting models,the combination model of BP neural network model and support vector machine model performed best and had the highest accuracy in freight demand forecasting.Through the study of this paper,the characteristics and influencing factors of freight demand in Tibet can be mastered.And a suitable method for the forecast of freight demand in Tibet can be constructed.The research results can not only provide reference for transportation planning in Tibet,but also provide reference advice for decision-making of relevant departments,which has certain theoretical significance and application value. |
