High-speed Train Load Factor Forecast And Breakeven Analysis

High-speed trains have become the main tool of railway passenger transport and the pillar of operating income, and economic benefit maximization is the core operation target of China Railway (CR) and railway bureaus. It’s urgent for transport enterprises to grasp the economic feasibility of train adjusting strategies or the decision to run a new train in advance. Profit and loss analysis of high-speed trains can provide a new perspective for transport organization optimization methods, which contributes to transforming the timetable into a cost-effective diagram. Therefore, it’s significant to make high-speed train profit and loss analysis.The paper defines that three components of train operating economic benefit are ticket revenue, transportation cost and value-added tax (VAT), and train load factor forecast is the research core of ticket revenue estimation. Firstly, it’s discussed whether BP neural network and multivariate nonparametric regression can be applied to forecast train load factor in different circumstances, based on factors analysis of train operation period, train individual characteristics and trains set with competitive relationship. And then according to train load factor prediction and the train average fare rate estimation, the ticket revenue estimation model is built. Secondly, with analysis of present research and application status at home and abroad, the transportation cost calculation model with activity based costing method is established and cost parameters are formulated. Thirdly, the VAT calculation method and rate are put forward. Lastly, train net profit and marginal profit are computed to analyze economic feasibility of adjustment strategies, and then profit and loss analysis is made from three levels of the EMU routing, EMU sub-routing and high-speed train to determine the scope of adjustment objects and the adjustment order.Taking high-speed trains of Beijing-Guangzhou high-speed railway in 2013 as an example, the overall research method is validated. Firstly, we analyze the train load factor prediction accuracy of MLP model and multivariate nonparametric regression model with the mixed data smoothing estimation method, and then confirm importance degree of the independent variables. Secondly, we analyze the annual average daily ticket income estimation error of 310 trains, which shows that the estimation method described in this paper is feasible. Lastly, the cost structure is analyzed on the basis of transport cost, VAT and net profit of 23 high-speed trains calculated, and then we make profit and loss analysis from three perspectives and seek out the scope of objects to adjust and adjustment priorities.