Lane capacity for an automated highway with mixed vehicle classes

This paper develops models to calculate an upper limit on per-lane throughp ut for an automated highway system with mixed vehicle classes and platooned operation. The models are analytical and based on independent arrivals amo ng the classes. The results indicate that a mixture of vehicle classes or c ategories can have a significant detrimental effect on vehicle capacity. Wi th a single entrance lane, capacity drops by more than 40% when the traffic stream moves from 100% light vehicles to a 90%/10% split between light and heavy vehicles. The capacity drop is substantially less when different cla sses are allowed to queue in separate lanes, but still significant (a 20% d rop). It also does not appear that sorting vehicles into destination groups at the point of entry is an effective strategy to enhance exit capacity. A separate model was developed to evaluate capacity losses at the point of e xit, where exiting vehicles must be separated by a minimum distance that ex ceeds normal intra-platoon requirements. Our results show that capacity los ses are moderate (10% or less) when less than 20% of vehicles exit at any p articular off-ramp.