Abstract:In the transportation planning literature, it is now well accepted that capturing uncertainty while evaluating transportation systems is important for arriving at better planning decisions. However, considering uncertainty presents numerous additional computational and theoretical challenges. This paper proposes an analytical methodology and efficient solution procedure that consider road capacity uncertainties as part of the overall transportation network model. The road capacity was considered as a random variable with certain distribution conditional to day-to-day roadway traffic conditions. For solution procedure, Sample-Average Approximation (SAA) was employed to generate plausible realizations of link capacity values from a multi-dimensional distribution and to solve the stochastic programming. INTRODUCTION There has been an increasing recognition of the need for effective approaches that incorporate the impact of uncertain events in transportation network models. It has long been recognized in the literature that capturing uncertainty in transportation system evaluation is important for arriving at better planning decisions (Mahmassani, 1984). However, planning by considering the effects of uncertain events, such as potential loss of vital transportation links as a result of incidents or accidents, adverse weather conditions and unfavorable road geometries such as steep climbs or lack of shoulders presents numerous theoretical and practical challenges. This study is concerned with the development of transportation planning models considering risk of uncertain events at the network level. The road capacity is considered as a random variable with a certain distribution conditional to the day-to-day roadway traffic conditions. The distribution of highway capacities due to random events were determined through the identification of historical accident frequencies for our network links from corresponding real-world databases and roadway capacity reductions from the literature. We then extended the Traffic Assignment Model with Capacity Uncertainty (TAMCU) for investigating the general travel time reliability and the impact of network link risk. The solution procedure for TAMCU relies on an approach referred to as Sample-Average Approximation (SAA), which has been extensively used to solve the stochastic programming problems, such as asset investment problem (Blomvall and Shapiro, 2007) and structure reliability analysis (Tsompanakis, Y. et al., 2008). PROPOSED METHODOLOGY AND APPROACH The traffic assignment has found significant application since Wardrop (1952) under the standard assumption of deterministic origin-destination demand and deterministic link capacity conditions. Typically, two types of traffic assignments are performed: User Equilibrium (UE) and System Optimum (SO). The formulation for each is given as follows (Sheffi, 1981):

Sampling-Based Approach for Incorporating Road Capacity Uncertainties into the Transportation Planning Process

Time-of-day Vehicle Mileage Fees for Congestion Mitigation and Revenue Generation: A Simulation-Based Optimization Method and Its Real-World Application

Modeling of urban road network traffic carrying capacity based on equivalent traffic flow

Capacity reliability of a road network: an assessment methodology and numerical results

Modeling Capacity at Signalized Intersections with a Left-Turn Storage Bay Considering Signal Timing Plan

Modeling Uncertainties for Automated and Connected Vehicles in Mixed Traffic

Tracing Road Network Bottleneck By Data Driven Approach

Capacity reliability of road network under adverse weather conditions based on fuzzy theory

Quantification and Reduction of Uncertainty in Seismic Resilience Assessment for a Roadway Network

Joint Service Capacity Planning and Dynamic Container Routing in Shipping Network with Uncertain Demands

A Distributionally Robust Optimization for Reliability-Based Lane Reservation and Route Design Under Uncertainty

Assessment of Road Network Capacity Reliability Based on the Constraints of Link LOS

Assignment of Freight Traffic in a Large-Scale Intermodal Network under Uncertainty

Monte Carlo Simulation In Sampling Techniques Of Traffic Data Collection

Robust Transportation Network Design under Demand Uncertainty

Reliable Routing of Road-Rail Intermodal Freight under Uncertainty

A Random Walk Approach for Simulation-Based Continuous Dynamic Traffic Assignment

A Safety-Aware Real-Time Air Traffic Flow Management Model Under Demand and Capacity Uncertainties

Practical Approach to Determining Traffic Congestion Propagation Boundary Due to Traffic Incidents

A Reliability-Based Stochastic Traffic Assignment Model for Signalized Traffic Network with Consideration of Link Travel Time Correlations

Modeling Traffic Networks Using Integrated Route and Link Data