The Impact Of The Autonomous Vehicles On Road Network Reserve Capacity

In order to study the impact of future autonomous vehicles on road network capacity,the vehicles on the road network were divided into two groups:autonomous vehicles and traditional vehicles.We assume that ordinary vehicles follow the user equilibrium and autonomous vehicles follow system optimum.The traffic demand multipliers among ODs adopt a unified growth mode;the reserve capacity model of the road network was constructed to satisfy the constraints of the road capacity.The results show that the road network capacity will increase with the enhancement of the market penetration of autonomous vehicles.On this basis,a non-uniform demand multiplier growth capacity model was constructed considering the OD demand structure on the impact of road network reserve capacity.And a multi-population genetic algorithm was applied to solve the non-uniform demand multiplier road network capacity model.The results show that when the market penetration of the autonomous vehicles is low,the growth of road network capacity is not obvious.When market penetration reaches a certain proportion,autonomous vehicles occupy a leading role.The road network capacity will increase approximately linear growth.Then,when the market penetration is large,the increasing trend of road network capacity slows down gradually.But the fluctuation is not too large.The growth trend of road network capacity calculated by non-uniform demand multiplier is similar to that calculated by uniform demand multiplier,but the former is larger than the latter.The growth multipliers of different OD pairs are not necessarily the same.The addition of autonomous vehicles can optimize the OD demand distribution in different areas,thereby enhancing the capacity of the whole road network.Considering the perception error of ordinary users on travel time,this paper assumes that ordinary vehicles follow stochastic user optimum and autonomous vehicles still follow system optimum.The road network capacity model was constructed under the mixed equilibrium flow.The results show that when the market penetration of autonomous vehicles and the familiarity of ordinary traveler users with network information are low,the capacity of road network increases rapidly.For a given market penetration,there is a critical value for the travelers' perceptual parameter,which corresponds to the maximum capacity level of the road network.Otherwise,the capacity of the road network will decline.When ordinary travelers are not familiar with the network information,there is little difference between the results of non-uniform and uniform demand multiplier.When ordinary travelers' familiarity with network information reaches a certain level,the centralized effect of network traffic distribution begins to appear.At this time,the network capacity calculation results of the non-uniform multiplier method are much better than the uniform multiplier method.According to the research,the calculation results of the network capacity that vehicles follow the user equilibrium are different from the stochastic equilibrium.And the latter is larger than the former in the autonomous environment.With the increase of the network travelers' familiarity with the road network,the difference of their calculations will gradually decrease.When travelers acquire the full road network travel information,ordinary users will follow the shortest path rule.At this time,the capacity of road network calculated by non-uniform demand multiplier is the lowest.So,the redundancy of network information is not conducive to the space-time equilibrium distribution of the road network traffic.