| With the rapid development of economy,limited road resources can hardly meet the traffic needs brought by social development.Therefor,drivers frequently change lanes in order to pursue larger sufficient driving space.Lane-changing,as a basic driving behavior,has a crucial impact on traffic flow.Reasonable lane-changing is of benefit to improve road capacity,while unreasonable lane-changing leads to higher traffic pressure.In this paper,the basic sections and diverging areas of expressway are taken as the research objects,and the cellular automata models corresponding to the vehicle operation characteristics in the study area are established respectively,which aim to explore the influence law of frequent lane-changing on traffic operation efficiency.Firstly,vehicle trajectory data of I-80 section in NGSIM(Next Generation Simulation Project)data set were selected in this paper.The symmetric exponential moving average method was used to smooth the data.Then,vehicle trajectory information was extracted before and after lane-changing,which was utilized to analyze vehicles’ lane-changing characteristics in different areas..Secondly,the pros and cons of the existing two-lane cellular automata models based on STCA two-lane model were demonstrated.Lane-changing probability function was optimized based on relative headway and relative speed.Considering the change of drivers’ driving behavior after being overtaken three times,the tendency function of driving behavior change was proposed.The lane-changing rules were improved by combining the above two functions.Moreover,a cellular automaton model considering discretionary frequent lane-changing was built based on the lane-changing rules.According to the three kinds of driving behavior characteristics of vehicles(default lanechanging on the left,symmetrical lane-changing,default lane-changing on the right)in the road system in the diversion area,the corresponding classical lane changing rules were selected as the research basis.Considering the distance between the vehicle and the latest lane change point,the probability function of mandatory lane-changing was optimized.Based on the above functions,lane-changing rules of vehicles was supposed to be improved,and a cellular automaton model considering mandatory frequent lanechanging was constructed.Finally,simulation scenes were built for tests.The simulation results of discretionary frequent lane-changing indicated that frequent lane-changing can promote the traffic operating efficiency under free flow,with the average speed and flow increasing by about5.89% and 4.16% at the maximum.On the contrary,it can inhibit the traffic operating efficiency under synchronous flow,with the average speed and flow decreasing by about9.48% and 14.91% at the maximum.The simulation results of mandatory frequent lanechanging indicated that: 1)when the on-ramp vehicle proportion was 0.05,the free flow speed increases by 9.7%,and when the off-ramp vehicle proportion was 0.1,the synchronous flow speed would decrease by 11.03%.2)the length of diverge area have both positive and negative effect on traffic.On the one hand,the increase of the length of the diverge area can provide a larger driving space for off-ramp vehicles,thus reducing the interference between vehicles and improving the average speed of the system.On the other hand,the maximum speed limit in the diverge area will drag down the average system speed.In the free flow state,the length of the diverge area had a greater inhibiting effect than a promoting effect on the traffic operation efficiency,and the opposite was true in the synchronous flow state.The research results can be applied to improve road facilities and regulate drivers’ behaviors.In the public sector,it can assist traffic management departments to formulate rules,regulations and management measures.In addition,this study contributes to provide theoretical foundation for vehicle collaborative system and automatic driving,which is conducive to promote the further development of driver assistance system. |
PDF Full Text Request