Research On Logistics Demand Forecasting In Chongqing Based On IAOA-LSSVM Model

With the continuous development of modern technology,the logistics industry has become one of the fastest growing new industries in China in the 21 st century.Chongqing,as one of the country’s top-ranked cities in terms of annual GDP,not only has a policy to make the city a regional hub for logistics in the southwest,but also occupies a unique advantage in terms of its geographical location,with its own Yangtze River basin,connecting the Asia-Europe Continental Bridge,and serving as a hub and gateway from East Asia to the Pacific and South Asia to the Indian Ocean.Chongqing’s flourishing logistics industry can promote the further development of the advantages of the western gateway hub,promote the process of urbanisation and industrialisation,and comprehensively enhance the overall comprehensive competitiveness of the city,which is of great strategic importance.It has effectively promoted the national strategy of western development.However,the overall logistics service capacity and level of Chongqing Municipality has not been able to meet its own development and industrial layout adjustment needs,which is inconsistent with Chongqing Municipality’s status and role as the economic centre of the upper reaches of the Yangtze River and the transportation hub of southwest China,as well as the development goal of achieving a national central city.Therefore,it is urgent and necessary to study the development of logistics demand in Chongqing.Based on the actual situation of logistics development in Chongqing and relevant theories,this thesis investigates the logistics demand forecasting in Chongqing through improved Archimedes algorithm and least squares support vector machine model.The research work mainly includes the following aspects:(1)Through relevant SPSS analysis,this thesis analyses the factors affecting the logistics demand in Chongqing city and explains the relationship between freight volume and Chongqing city economy.Combined with the current situation of logistics development in Chongqing,the main factors influencing the logistics demand in Chongqing are studied and a prediction index system for logistics demand in Chongqing is constructed.The indicator system is simplified by quantitative analysis of indicators through entropy weight method and grey correlation analysis.The influencing factors with high correlation with the logistics demand in Chongqing are selected to forecast the logistics demand in Chongqing,laying the foundation for the research of the city’s logistics demand forecast.(2)The logistics system is a complex non-linear system characterised by small data samples and a lack of intrinsic connections and regularities.To address these problems,this thesis finds that least squares support vector machines(LSSVM)have unique advantages in solving the problems of finite samples,non-linear functions and multidimensional model identification by comparing the advantages and disadvantages of various forecasting methods.Therefore,in this thesis,the LSSVM model is applied to the study of logistics demand forecasting in Chongqing city.Firstly,we chose the RBF kernel function with high classification accuracy.Then,an improved Archimedean algorithm(IAOA)is designed in this thesis,which is mainly improved by SPM chaotic mapping,modification of the density update formula and introduction of Fevy flight.Finally,a comparative performance analysis of the improved algorithm is carried out using five algorithms and six functions.The results show that the algorithm not only improves the optimization search capability,but also increases the iteration speed.(3)In addition,this thesis also analyzes the forecast results of Chongqing City from2002 to 2021 by comparison,and the research results show that the logistics demand in Chongqing City shows a trend of year-on-year growth,and the growth rate is gradually accelerated.This conclusion has a reference value for the development of the logistics industry in Chongqing,and can provide a scientific basis for relevant policy formulation and enterprise decision-making.Therefore,the optimised LSSVM forecasting model based on the IAOA algorithm in this thesis has high forecasting accuracy and adaptability in forecasting the logistics demand in Chongqing,and can provide scientific forecasting and decision-making support for the development of the logistics industry in Chongqing.