This paper describes a project to prepare a proposal, whereby the Maine Department of Transportation (MDOT) would comply with the requirements of the USA's 1990 Clean Air Act. The project's Phase 1 was to establish 1990 baseline levels of traffic congestion, as measured by total vehicle delay at 372 signalised intersections. Phase 2 was to recommend and quantify traffic type improvements, designed to reduce congestion and decrease total vehicle delay by 15%. The following data were collected: (1) traffic turning movement counts, at 15min intervals, at 335 intersections; (2) 24-hour traffic volume counts, at 15min intervals; (3) existing geometric conditions at each intersection; and (4) existing traffic (4) existing traffic signal equipment. Total and approach vehicular delays were analysed. In Phase 2, all signals were analysed for all hours of normal operation, and all signal systems were analysed using TRAF-NETSIM, a stochastic, microscopic traffic simulation model for urban street networks. 28 signal systems for 127 intersections were investigated. Challenges faced included: (1) the large data collection effort; and (2) the many calculations required to estimate vehicular delay. The paper shows how TRAF-NETSIM was used to address them.
Abstract