Development of a car-following model to simulate driver and autonomous intelligent cruise controlled vehicular traffic flow

The objectives of this research were to; (1) examine the existing microscopic simulation models to reveal their assumptions and deficiencies, (2) develop a car-following model to simulate driver car-following behavior more realistically than the existing models and (3) study the effects of Autonomous Intelligent Cruise Controlled (AICC) vehicles on traffic capacity and flow.; The car-following models NETSIM, INTRAS, FRESIM and CARSIM do not realistically simulate the driver car-following behavior and these models have some deficiencies. In order to overcome these deficiencies, INTELSIM car-following model has been developed. INTELSIM moves vehicles simultaneously and employs a linear acceleration model to simulate the drivers' car-following, stopping and accelerating to their desired speeds behavior. Driver perception thresholds are considered and the preferred time headways (t_p) which represent the separation headway during steady state car-following are utilized. The driver reaction times are applied realistically and are independent of the time steps. The successive reaction times, named as chain reaction times are introduced. The vehicles in car-following reach their steady states gradually, not in one time step. The drivers' perception thresholds, reaction times and preferred time headways result in a behavior that closely resembles the real car-following behavior. INTELSIM model has been validated with different sets of car-following data. Statistical methods to be used for the validation of simulation models are explored.; The performance of the analytical and simulation car-following models were assessed. It was observed that the generalized car-following (GCF) model parameters that satisfy macroscopic relationships do not always yield appropriate car-following behavior. The GCF model was not able to simulate driver reaction times.; INTELSIM model was also utilized to study traffic flow characteristics of a mix of driver controlled (DC) and AICC vehicles as well as all AICC vehicles. INTELSIM algorithm provided stability in AICC platoons and performed well in a variety of traffic situations including stop-and-go conditions. AICC vehicles can provide volumes of up to 7400, 5200 and 3100 veh/hr at separation headways of 0.3, 0.5 and 1 sec, respectively, at speeds of 90 ft/sec. These volumes are much greater than those obtained by DC cars. The traffic density and volumes of traffic mixed with AICC and DC vehicles were also examined. Macroscopic traffic flow equations to represent AICC vehicular traffic were found.