Unsignalized intersections are often modeled with an assumption of random vehicle arrivals. Vehicle arrivals are not random where nearby upstream signalized intersections cause platoons to form. A method is presented to identify periods of high and low flow. A statistical procedure is developed to estimate the high and low traffic flow rates caused by an upstream signal. The procedure is demonstrated with data from unsignalized intersections affected by platoon flow. Ramifications of using the procedure in the analysis of unsignalized intersections are then discussed.
Abstract