Research On Robust Optimization Method Of Public Transport System In Random Environment

Public transport is a complex and open system,relevant with the urban traffic system and the urban residents travel,that various uncertain factors are likely to change the steady operate state of public transport system.With the rapid development of social economy,as one of the most important way to travel of urban residents,people pay more attentions to the reliability and stability of public transport system,that it is referred to as "robustness" of the bus system in this paper mainly embodied in the terms of running state of single bus route.Urban traffic system contains a large number of random disturbances.Especially,travel speed and the uncertainty and randomness of bus passengers arrived at the bus stops will be impact on public transportation system seriously,leading the buses in the running process influenced by traffic disturbance difficult to accurately follow the settled operation scheduling,and the bus system failing to maintain the stability of the operation.In view of the disorderly bus operation problems induced by random disturbances,focus on the problem of bus bunching in bus system,analyze the traffic flow characteristics influenced by the bunched two buses and build the vehicle additional delay calculation model based on traffic wave theory.Considered the stochastic volatility of passengers arriving at bus stops,based on the statistics and analysis of two dimensional time series on bus passenger flow data,design a two-dimensional time sequence prediction model of passenger flow at bus stop.Using the means of bus holding and coordinated control,and the concept of self-adaptive control strategy to equalize headways,propose an robust optimization method for bus system operation under random environment,improving ability to resist the random disturbance of public transport system.According to discussions and researches on the major steps of public transport system control method,form a systematic and scientific bus operating real-time control method system.The main results and work of this research are summarized as follows.(1)Additional cars delays calculation model based on shockwave theory.In this section,the problem of traffic flow influenced by bunched buses is studied,while the bus stop is located near the signalized intersection.First,based on the analysis of the impact on the traffic flow by single bus at bus stop,the cars delays calculation is divided into six kinds of situations.Based on the shockwave theory,the additional delay calculation model under the influence of single bus is formulated,to obtain the delay of cars on the street under the influence of single bus in different circumstances.By the relationship between the additional cars delays and the bus dwell time,analyze the operation mechanism of the two bunched buses under the one berth and two berths situations,and design the total dwell time of the two buses at bus stop to obtain the relationship between dwell time and additional cars delays,and formulate the calculation model of additional cars delays under different berths.The research results in this section can provide theoretical support for the necessity of the implementation of real time control method for public transport.(2)Short-term bus passenger flow forecasting method based on two-dimensional time series data.Passenger flow volatility of arriving at bus stops is the main random factor considered in this paper,which influences the stability of public transport system.Considering the predictive demand of short-term bus passenger flow data,the two-dimensional time series passenger flow data(real-time dynamic passenger flow data in current day and historical passenger flow data)are taken as the basic data,and the Kalman filter algorithm is used as the foundation.Use parallel forecasting architecture and decision-level fusion method to design the bus passenger flow forecasting procedure based on the data of two-dimensional time series.With the way of dynamic prediction of adaptive weights and the form of rolling prediction method,considering the simplicity and precise of the predictive model,design three bus passenger flow decision-level fusion models,providing suitable passenger flow forecasting methods for different situations.The research in this section provides accurate data support for the robust optimization of public transport system.(3)Robust optimization method of public transport based on bus holding.Based on the description of the bus operation process in the public transport system and the identification of main stochastic factors affecting the bus operation,the mechanism of the bus bunching in the public transportation system is analyzed.The bus system is abstracted and public transport system is described with mathematical linguistics.With the point of view of improving convergence rate of buses headways and reducing the holding time at control points,and the concept of self-adaptive to equalize buses headways,propose a control strategy making the successive unbalanced buses at control point recovery.Because of the random disturbance factors in the public transport system,it is difficult to formulate the analysis control model based on bus holding.In order to overcome the problem,from the simple to complex,from the ideal situation to the actual situation,formulate bus holding time calculation model and robust optimation model based on bus holdig control.The proposed control method is tested with three parts,which are the theoretical derivation under ideal situation,numerical analysis under the linear and non-linear assumptions and the simulation under the actual situation.The results show that the proposed control method can improve the ability of public transport to resist random perturbation.The research in this section provides a new method to improve the ability of public transport system against random disturbance.(4)Robust optimization method of public transport based on coordinated control.Considered the adaptability of robust optimization method for public transport system under different passenger flow demand,based on the analysis of the advantages and disadvantages of skip-stop and bus holding,this paper analyzes the benefit bring by the coordinated control method including the two control means.With combination of dynamic skip-stop and bus holding,and the control concept of self-adaptive to equalize buses headways,select the two control means dynamically according to the headways at control point,formulating a coordinated control method procedure.Describe the evolution process of buses headways in public transport system under different control means,and dynamically select the proper control mean according to different control results,calculating the holding time if the selected control means is bus holding,and put forward the robustness optimization method of public transportation system based on coordinated control of bus holding and sip-stop.The dynamic coordination control method is used to analyze the selection of various control strategies in the coordinated control system by the means of the numerical analysis method,according to the selection of each control method.The research in this section provides the scientific control method for the construction of intelligent control system of public transport system.Considering the applicability,simplicity and advancement of the proposed methods,to solve the bus bunching problem in urban public transport system,propose robustness optimization methods in random environment,using the advanced scientific methods in the urban public transport system application.The research thought,context and the method are the supplement of the urban bus operation and management system.The research results have important academic value.