Determining optimal sensor locations in freeway using genetic algorithm-based optimization

Travel time is the most intuitive measure of effectiveness for road users and transportation agency operators. However, travel times derived from speed data measured at fixed point sensors often varies from actual travel time. This is, in part, due to the intentional positioning of sensors to avoid lane changing and/or to inadequate numbers of sensors capturing the dynamic characteristics inherent in freeway traffic flow. This paper presents an approach that optimizes the location of sensors in a freeway to support more accurate estimations of travel times than those obtained from conventionally deployed fixed point sensors. Evaluation results, under varying traffic conditions, including incidents, showed that the proposed approach produced average travel time estimation errors within 10% and performed much better than the conventional approach. Thus, the proposed approach provides a promising tool to support re-positioning of the existing non-intrusive point sensors (e.g., video sensors) or deployment of new sets of point sensors for improving travel time estimation.