Optimization Models Of Line Planning For Passenger Dedicated Lines

Line planning plays a key role in railway transport system, and it is an essen-tial bridge between passenger flow and train diagram. An excellent line planning may increase the operating income of railway companies and enhance its competitiveness. Establishment of line planning in our country more depends on the experience of staff, which is not only time-consuming, but also lack of quality. Therefore, to establish line planning systematically and automatically is an important problem to be solved urgently.In this paper, three optimization models of line planning are established, considering both the railway companies and passengers. These models are applied to the line planning design of Shanghai-Nanjing High-Speed Rail.The optimization model, minimizing the operation cost of periodic line planning, is established to meet the demand for passenger travel. This model assumes that the train stops at every station. A series of important inequalities are proved. By calculation, it is found that the transfer case may be occurred, this will not only increase the passenger travel time, but also miss the subsequent train. So, an optimization model without transfer is established.Besides, passengers have initiative to choose trains. Both the railway company’s operating income and the passengers’satisfaction should be fully considered. Reducing the waiting time can increase passengers’satisfaction, so based on the transfer-free model, an optimization model, minimizing the waiting time, is established. This model can obtain the optimal schedule of the train and the corresponding passenger flow assignment, which have important reference value to the design of ticket allotment plan.Moreover, all-day passenger flow has obvious bimodal characteristics, so the trans-port capacity at rush hour should be larger than at other time. In order to meet pas-sengers’ demand as possible, the concept of "online train number" is introduced, and a nonlinear integer programming model of aperiodic line planning is established. By in-troducing several decision variables, it is transformed into a linear integer programming model, which greatly reduces the difficulty of solving this problem.