A Space Discretization Based Simulation Approach for Non-lane Based Traffic Conditions

Typical non-lane based traffic is characterized by the lack of lane discipline and the presence of multiple vehicle types. Moreover, mixed traffic streams require modeling of continuous lateral movements which are only restricted by the road width. The existing simulation tools may not be able to replicate these specific characteristics as most of them are developed based on the underlying concepts of lane based driver behavior. To address these critical trade-offs’ between the lane-oriented and the space-oriented movement behaviors, a new space discretization based simulation approach is proposed. In this approach, the lane is divided into small bands or strips and the vehicular movement is tracked along them. In doing so, the dynamic virtual lane oriented movement behavior as seen in mixed traffic environment is achieved. However, such a concept suggests the movements of vehicles as being still lane-based in which they are merely regarded as a small/large car that travels in sub-lanes. So, this description does not cover the typical longitudinal and lateral movement behavior of different vehicle types precisely. To address this critical issue, behaviorally better longitudinal and lateral movement models are proposed. The proposed longitudinal movement model incorporates the leader-follower combination based following behavior. This is done by converting the gap, a specific leader-follower requires into an equivalent gap maintained between car and a car. For this, regression equations are developed correlating the gaps of car-car combination and the other vehicle type combinations. The lateral movement model calculates the gain in making a lateral shift in anticipation of better driving conditions using a new benefit function. These models, when incorporated into the proposed space discretization based simulator, work behaviorally better than the pure lane-based models as they allow the tactical following and overtaking maneuver by a driver along the dynamic virtual lanes as seen in mixed traffic scenario. Moreover, the proposed simulation approach retains the concept of lane based movement when the strip width equals the traditional lane width and so can be used even for studying the traditional lane-based traffic streams. The model was calibrated and validated with data from Mumbai, India and the results indicate better representation of the mixed traffic movement.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; Photos; References; Tables;
  • Pagination: 12p
  • Monograph Title: TRB 93rd Annual Meeting Compendium of Papers

Subject/Index Terms

Filing Info

  • Accession Number: 01516094
  • Record Type: Publication
  • Report/Paper Numbers: 14-3854
  • Files: TRIS, TRB, ATRI
  • Created Date: Jan 27 2014 3:19PM