Application Of Improved Grey Neural Network Model In Gansu Railway Passenger Volume Forecast

For a long time,railway transportation has been the lifeblood of economic growth,but also one of the main way people travel.In recent years,with the continuous improvement of living standards,people's travel demand increases and stimulates the growth of railway passenger transport.But the urgent,as the basis for the construction of rail passenger,need to accelerate the construction of the railway sector and traffic forecast is playing an increasingly important role.It can be seen that it is necessary to make scientific and reasonable forecasting of railway passenger traffic.It can not only provide the basis for railway transportation industry,but also provide a reliable index for railway evaluation system.At the same time,passenger traffic volume is affected by many factors.With the great development of the western region and the sustained growth of the economy in Gansu Province,the railway passenger traffic volume of Gansu province is on the rise.With the above background,this paper first describes the background,purpose and significance of the research on the prediction of railway passenger volume,analyzes the research status of domestic and foreign railway passenger volume forecast,based on the introduction of the main contents of this paper and the technical route.Secondly,this paper expounds the commonly used prediction method in the research of railway passenger capacity,and with the Gansu province railway passenger volume data and passenger volume trend,established time series based on Grey GM(1,1)model and grey Verhulst model forecast.Again,from the qualitative point of view,expounds the relationship between railway passenger volume and its influence factors,and the development trend of the passenger volume and influencing factors of change will influence factors were quantified,using gray correlation analysis method to calculate the correlation degree between each factor and the passenger volume,to determine the main influencing factors railway passenger volume,established BP neural network model for prediction.Then,through the analysis of the advantages and disadvantages of the grey prediction model and the BP neural network model,the grey BP neural network combination model is constructed,and the model is used to predict the passenger traffic volume.Finally,this paper compares the prediction results of the above models,and finds that the average relative error of the combination forecasting model is minimum.Moreover,it is found that the model residuals satisfy non-stationary stochastic properties and exhibit Markov properties.Therefore,the Markov chain model is used to optimize and improve the predictive value of the grey BP neural network combination model,and further improve the model accuracy.Based on the study of passenger volume forecasting,a grey BP neural network combination model based on Markov chain is proposed,and the prediction accuracy of this model is relatively high by the example analysis.It can provide important guidance for the railway construction and transportation planning of Gansu Province,and provide a new solution for the forecast of passenger traffic volume.