Research Of Mixed Traffic System With Autonomous-manual Vehicles Considering Driving Behavior

The emergence of rudimentary autonomous vehicles providing a new way to ease the road traffic congestion problem.However,there are many uncertainties about the impact of autonomous vehicles on traditional road traffic systems as they gradually enter the market in the future.In order to fully understand the characteristics of the mixed traffic flow formed by autonomous-manual vehicles,two kinds of automatic-manual hybrid traffic flow cellular automata models of one lane and three lanes were established in this study,which were fully considering the driving behaviors of these two different vehicles.Before modeling,we proposed three hypotheses to distinguish autonomous and manual vehicles:(a)Autonomous vehicles won’t be effect by random moderating,while manual vehicles will be effect.And human driver’s mentalities are subject to change with the increase of the traffic density.Usually,the greater the psychological pressure is,the bigger the random probability of slowdown will be.(b)The reaction time of autonomous vehicles is smaller than manual vehicles.And the more experienced the driver is,the shorter the reaction time will be.(c)Autonomous vehicles can accurately estimate the speed of vehicles around,while manual vehicles can’t.Here we used driving-style coefficient to describe the ability to judge the speed of other vehicles.According to Hypothesis(b)and(c),we proposed the "Expected Safety Distance Model" based on GIPPS safety distance rules.And combined with Hypothesis(a),we established “Single-lane Autonomous-manual Driving Mixed Traffic Flow Model”.Then we improved the lane change rules by the proposed "Critical Lane-change Probability Model" together with the "Expected Safety Distance Model" and established the“Triple-lane Model”.Lastly,we used MATLAB R2016 a to carry out the mixed traffic flow simulation under different scenarios.Through the research we found that: The maximum flow of the single lane system can be promoted 1.2 times,triple-lane system can be promoted 1.4 times.Besides,under the scenario of the dedicated lane,when the market penetration of autonomous vehicles reaches20%,there is little difference between triple-lane system mixed traffic flow and the dedicated lane system,thus,considering the safety of the autonomous vehicle,dedicated lane could be established at this market penetration;And when the market penetration of autonomous vehicles reaches 35%,the capacity of the triple-lane system will be the highest.