Font Size: a A A

Research And Application Of Bus-only Lane Scheduling Optimization Algorithm

Posted on:2024-06-30Degree:MasterType:Thesis
Country:ChinaCandidate:A N DuFull Text:PDF
GTID:2542307079960539Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Urban public transportation has many advantages such as high efficiency and low pollution,and plays a key role in alleviating urban traffic congestion and reducing urban pollution.The establishment of bus-only lanes is one of the key measures to achieve this goal.The scheduling optimization of bus-only lanes can be refined into the scheduling optimization of vehicles on bus-only lanes and the signal light scheduling optimization at both ends of bus-only lanes.This thesis focuses on two optimization methods and carries out technical research and application on road traffic flow model for bus-only lanes,effectiveness and setting conditions of bus-only lanes,and signal control of bus-only lanes.The main innovations and work of this thesis are as follows:(1)For the vehicle scheduling optimization problem on bus-only lanes,a road traffic flow model for bus-only lanes is proposed.This model uses an improved cellular automaton model to model road traffic flow for bus-only lanes,and considers the behavior models of different vehicle types on different lanes,proposes a bus priority index to reduce the impact of social vehicles on buses when changing lanes to enter bus-only lanes.This thesis conducts simulation comparison experiments under three bus-only lane setting methods.The experimental results show that the road traffic flow model for bus-only lanes proposed in this thesis can improve road traffic volume and passenger volume and alleviate road congestion.(2)For the signal light scheduling optimization problem at both ends of the busonly lane,a signal control model for bus-only lanes is proposed.This model uses a deep reinforcement learning algorithm,proposes the concept of multi-priority road rights,and applies it to the proposed signal control model,making it applicable to traffic conditions with multiple priority vehicles.This thesis conducts a comparative simulation experiment with a timed signal control scheme.The experimental results show that the signal control model for bus-only lanes can ensure the priority passing ability of buses at intersection sections and reduce the average delay time of vehicles.(3)A traffic simulation system for bus-only lanes is designed and implemented.This system adopts a hierarchical architecture,including a traffic simulation front-end page,a traffic simulation system backend,and a traffic simulation engine from top to bottom.The traffic simulation engine is the key part of the system,which implements core functions related to traffic simulation,supports adding buses and bus-only lanes in the road network,as well as other functions such as simulation result analysis,and verifies the availability of system functions through system testing.The core work of this thesis includes theoretical analysis and practical application.The road traffic flow model and signal control model help to achieve optimized scheduling of bus-only lanes; the traffic simulation system for bus-only lanes can simulate,verify and evaluate bus-only lane strategies.The research results have wide application value in many fields such as bus network planning and urban road design.
Keywords/Search Tags:Bus-only Lane Scheduling, Vehicle Behavior Model, Cellular Automata, Reinforcement Learning, Traffic Simulation
PDF Full Text Request
Related items