Research On Evaluation And Adjustment Of High-speed Railway Line Plan Based On Dynamic Demand

Accompanying more high-speed rails put into use and more gradually added to its network, the passenger transportation problem has changed from just space transfer to a service problem satisfying various travel demands.The needs for convenience,express,comfort and other quality requirements are improved gradually.And, passenger flow demand regularly surges with the variation of operation periods.This mainly reflects large differences of passenger flow volume, origin-destination and flow structure between different operation periods(daily operations,holiday operations).The train operation plan which was decided based on passenger flow forcasting before operating cannot adjust to dynamical changes of passenger flow well.According to various travel demands of opertion periods, adaptability evaluation and adjustment of train operation plans are put forward, being of great value in perfecting passenger product stucture, improving passenger service quality, developing products and maintaining competitive advantages.The adaptability evaluation and adjustment of train operation plans based on dynamic passenger flow demands are explorations on theories and methods of railway passenger transport organization. Its mainly content includes that by introducing concepts of train operation plans and flow demands firstly, a correlation between them is analyzed. According to flow demands of each period and adaptability of operation plans, the problem of whether or not adusting plans is analysed and then the adjustment modle is presented as well. And quantified as well as their adaptability evaluation indicator system and evaluation method are set up;based on dynamic flow demands and operation plans, a model of passenger flow-assignment of railway transportation networks is proposed, and methods of calculating indicators are put forword based on the results of assignment and needs;for operation plans cannot adjust to dynamical changes of passenger flow well, ajusting strategies of train operation plan based on operation period partitioning are presented.The achievement of the paper is as follows.1. The characteristics of high speed railway train operation plan and the demand of passenger flow are systematic analyzed, and the concept and content of the two concepts are clearly defined and deepened. The correlation between high speed railway train operation plan and the demand of passenger are analyzed from interaction mechanism. This paper describes the mechanism of passenger flow demand on train operation plan from the aspects of passenger flow to train operation plan preparation and the adjustment of train operation plan with the dynamic change of the passenger flow, expounds the mechanism of the train plan on the passenger flow demand in terms of the railway passenger's choice behavior under the established train plan. Temporal dynamics of passenger flow demand of high-speed railway is analyzed in perspective of daily and annual passenger flow demand. Given that it's hard to regulate train operation plan according to real-time passenger flow fluctuation, it puts forward the train operation plan based on the classification of passenger flow period.2. The affinity propagation clustering algorithm is employed to take dayly arrived and departed passengers of station as study samples, make clustering of the affinity time point(day) in passenger statistical data, and defines the best clustering number with indicators of Calinski-Harabasz, Hartigan and In-Group Proportion in order to achieve the annual operation period division for high-speed railway.3. The train operation plan and passenger flow demand adaptability analytical indicators system for high-speed railway is set up based on three aspects: capacity adaptability,structure adaptability and quality adaptability. The passenger flow demand satisfaction rate and average occupancy are employed to evaluate the capacity adaptability of both aspects.Structure adaptability is measured by the matching between passenger transportation quality and quantity in perspective of each passenger flow demand. And transportation supply-demand service quality adaptability is measured by the crowd of train and passenger transfer rate. Focus on the redundancy appeared in the index system, knowledge reduction method based on rough set was adopted.4. Based on the analysis on inadequate real-time monitoring data acquaintance method of the ticketing and reservation system, it puts forward the adaptability evaluation indicators calculation method based on flow distribution stimulation. It analyzes the influential factors on railway passenger traveling plan selection, depicts the generalized travel expense of individual passenger's selection in perspective of price expense,travel time and train crowd,employs service arc section ability and passenger transfer times as restriction conditions,establishes the train service network passenger distribution model which can minimize passenger's generalized expense, and proposes the mixed passenger distribution iterative algorithm based on improved Dijkstra and MSA. Eventually, it offers precise analytic data for train operation plan and passenger flow demand adaptability indicators calculation. In addition, the method integrated of information entropy and matter element analysis is employed to determine the weight of each indicator and the final evaluation results of train operation plan and passenger flow demand adaptability.5. Four elements of train operation plan: operation segment, stop plan, operation frequency and train preparation mode are transformed into adjustable decision-making variants. The definition of objective function takes the maximum passenger flow demand satisfaction, passenger flow consuming time and minimum train operation plan cost at all classes, as well as the high-speed railway train service quality and stop number restriction in consideration. On such base, it sets up a dynamic passenger flow demand-based railway high-speed train operation plan adjustment and optimization model, and designs the stimulation annealing algorithm according to model feature to get the results.