Research On The Influence Of Weather Factors On Passenger Flow Of High-speed Railway

Railway passenger flow prediction is one of the important work of the railway passenger transport departments to understand change of passenger flow and grasp the market information,the short-term passenger flow forecasting is the basis of railway passenger transport departments to deal with market change in timely through short-term capacity adjustment,ticketing organization optimization.However,the short-term passenger flow influenced by many factors,the weather is one of the important factors,it has a promoting effect to improve prediction accuracy,extend the applicable scope of models while short-term passenger flow prediction to consider the factor of weather.For this reason,firstly,this paper analyzes the characteristics of random fluctuation,interannual fluctuation,weekly fluctuation and holiday fluctuation of passenger flow data of Beijing-Shanghai high-speed railway,and integrates and analyzes with weather data to calculate the passenger flow reference value and passenger flow deviation rate,so as to provide data basis for studying the influence of weather on high-speed railway passenger flow.For some ODs of Beijing-Shanghai high-speed railway,discusses the rainfall weather factors such as wind and air temperature,air quality,the impact on the high-speed rail passenger flow mode,impact strength.on this basis,this paper indicates the rectification method of high-speed rail passenger flow prediction with basis of passenger flow prediction using conventional method and consider impact of weather and build a high-speed rail passenger flow forecast correction model with considering the influences of the weather.Finally,paper select the history of the Beijing-Shanghai high-speed rail passenger flow data,using nonlinear regression prediction model of SVR to predict short-term passenger flow,then using the recycling correction model while considering the influences of the weather to modified the forecasting results,and compares the prediction errors of before and after correction,to verify the effectiveness of the proposed method and model.44 figures,20 tables,43 references.