An Exploratory Study of Potential Freeway Level-of-Service Measures by Simulation Approach and Statistical Analysis

Abstract

Traffic operations analysis for freeway management includes Level of Service (LOS) methodology. Conventionally, density is adopted as the LOS measure for basic freeway segment. To account for different vehicle types in mixed freeway flows, the current method transforms trucks into passenger car equivalent (PCE) values. This practice has raised questions regarding its effectiveness because drivers of different vehicle types perceive the quality of service differently even under a same density. Therefore, "whether the PCE-based adjustment for a single LOS threshold for all vehicle types is reasonable" is subjected to investigation. So this paper proposes two measures and evaluates their applicability to major vehicle types: acceleration noise (AN); average lane change rate (ALCR). Their respective relationship to traffic and operational factors was investigated through simulation approach, and totally 108 simulation scenarios resulted from the factorial design. AN showed promising results as it exhibited differentiable monotonous tendency, with the flow rate changing, for each of different vehicle types, while ALCR showed the vulnerability, namely the lack of monotonous and differentiable tendency, regarding the service quality measurement for different vehicle types. This study provides some inspiring insights into further researching and developing freeway LOS measures for different vehicle types in burgeoning big-data environments.