Research On Port Throughput Prediction Of Tianjin Port Based On PCA-SVR In The New Era

With the implementation of the "One Belt and One Road" and the Beijing-Tianjin-Hebei integration policy,as the important trading port of the Beijing-Tianjin-Hebei integrated urban circle and an important strategic fulcrum of the "One Belt and One Road" policy,Tianjin port vigorously develops foreign trade and strengthens the trade with the countries and regions along the route;which further improves the Tianjin Port collection and distribution network system,enhances the radiation capacity of the port,and consolidates the inherent advantages of Tianjin Port as the northern international shipping core area.As an important indicator to measure the development of port logistics,predicting port throughput reasonably and mastering its changing laws are of great significance to port layout,berth location,development strategy and business strategy.Firstly,this dissertation analyzed the factors affecting the cargo throughput of Tianjin Port under the "One Belt and One Road" and the Beijing-Tianjin-Hebei integration policy,and the port infrastructure,the collection and distribution system,the economic development level of the hinterland,the level of foreign trade development and the background of the times were mainly considered.From that,the influencing factors index was established to analyze the impact of Tianjin Port’s own construction,hinterland economy and the background of the times on port throughput.Secondly,the principal component analysis was used to conduct a principal component analysis on the nine selected index variables.Four main components with a cumulative contribution rate of over 99% were extracted as the key factors affecting the cargo throughput of Tianjin Port,and they were used as the input data of the model.Thirdly,in order to improve the prediction accuracy,this dissertation used the support vector regression with strong nonlinear modeling ability to predict the throughput of Tianjin Port,and compared the PCA-SVR combined prediction model with the linear regression model and the random forest model through empirical analysis.The results show that the PCA-SCR combined prediction model has higher prediction accuracy and better effect,which provides new research ideas and methods for port cargo throughput prediction.Finally,the constructed PCA-SVR combined forecasting model was used to predict the throughput of Tianjin Port from 2019 to 2025.Based on the forecast results,the throughput growth rate of Tianjin Port and its influencing factors were analyzed.The research shows that the growth rate of Tianjin Port throughput presents a downward trend in the next few years.And in numerous influencing factors,the three indicators of the total value of imports and exports of foreign trade commodities,the added value of tertiary industry and Hebei port throughput accounts for the percentage of total throughput of Beijing-Tianjin-Hebei port have a significant impact on Tianjin Port throughput.