| Urban road traffic is an open and complicated system that is comprised of many elements such as vehicle flow,passenger flow,network infrastructure,etc..The mixed vehicle flow consisted of cars and buses is the most important part of road traffic.From the perspective of vehicle flow system,it is a basic characteristic of urban road traffic that cars mix with buses.Considering cars’ large proportion in the urban road traffic,their traffic state can be regarded as the external environment of bus system,and car flow has a significant impact on the operational efficiency of buses,which may affect the implementation of bus dispatching.However,to control the complexity of the problem,traditional bus dispatch management usually simplifies this impact by using random variables or even ignoring the car flow.Thus,it has limited effect under the open circumstances of urban road traffic when application.Consequently,it is necessary to integrate car flow control with bus dispatching for the transforming and upgrading of bus system,as well as improving of the total network efficiency.This thesis shifts from the traditional perspective of “microscopic bus route dispatching” and “single bus system” to the perspective of “macroscopic traffic regional control” and “mixing of buses and cars”,thereby exploring new idea of traffic management with large-scale network on the premise of controlling the complexity of the problem reasonably.Furthermore,considering the large occupancies of buses,it is not fair to optimize vehicle flow when road resources are limited.Thus,this thesis proposes a macroscopic perimeter control integrated with bus dispatching based on multi-modal passenger MFD(Macroscopic Fundamental Diagram)to maximize the total passenger arrival rate.Firstly,the type-II fuzzy logic is adopted to establish the travel cost estimation model of each mode from two inputs to one output.The travel cost of buses is calculated based on the average travel time and the crowding degree of buses;And travel cost of cars comes from the average travel time and the extra cost(owning to the car flow control)of cars.According to the travel cost of each mode,the classic logit model is used to transform regional passenger OD to regional vehicle OD,which relates to choice behavior for different traffic modes.Secondly,based on the range of optimal ratio of bus accumulation to car accumulation which is derived from multi-modal passenger MFD,the rate of car entering the specified region is adjusted by setting dynamic release rate to provide external guarantee of bus system;Regional frequency is defined as the number of bus departures per unit time in the specified region to control the bus accumulation,thus improving the efficiency of bus operation and reducing the extra cost of cars due to perimeter control.Finally,real data collected from the part of Futian District in Shenzhen are utilized for numerical simulation by Matlab to verify the efficiency of the proposed control strategy.The results of the simulation experiments conclude that:(1)The proposed integrated traffic control strategy can maintain high efficiency of the given network under the condition of high passenger demand;(2)There is relatively large limitation on cars whose demand is external,but the release rate at the border is stable which has good guidance for signal control in future work;(3)The proposed control strategy can reduce the average occupancy of buses properly according to the comparison of different control strategies,which is able to lower the crowding degree and to improve the attraction of buses. |
