| A car following model, CARSIM, is developed and validated under various traffic conditions at microscopic and macroscopic levels. It realistically simulates normal flow as well as stop-and-go operations. CARSIM provides a marginally safe following distance for all vehicles. It uses variable reaction times for drivers, and takes into account the start-up delay of the drivers. CARSIM may be used as a single regime model or as a dual regime model.; After an extensive validation the model is used to study the propagation of traffic waves at a microscopic level and the effects of mainline traffic control on congestion alleviation. The effects of the lead car's deceleration rates and traffic volume on the propagation of slowdown and stopping waves are investigated. A set of equations and graphs are developed to find the queue length, speed of wave propagation, location of a wave, time, and the number of vehicles affected by the waves. The rates the vehicles will be affected by the slowdown waves are 2.75, 2.13, and 1.71 sec/vehicle, and the speed of propagation of the slowdown waves are -9.1, -11.8, and -14.8 ft/sec; and that of the stopping waves are -9.1, -11.74, -14.53 at traffic volumes of 1200, 1500, and 1800 vph, respectively.; The effects of an incident the propagation and dissipation of a slowdown, stopping, starting, and recovering waves are investigated. The slopes of the waves are -9.41, -9.17, -13.51, and -12.38 ft/sec., respectively. The effects of a 3 minute incident lasted about 10 minutes. A strategy which guides the vehicles before reaching the bottleneck is compared with not-guiding. The average number of stopped vehicles decreased from 186.2 to 114.0 (26%). The average time a vehicle spent to travel the section is reduced in spite of a deliberate delay that is caused to upstream traffic by asking them to slowdown. When the traffic was guided the congestion is eliminated 141 seconds faster. |
