Transit Network Design is a complex problem, traditionally addressed with optimization models based on fixed demand assumptions. A new study introduces an innovative framework, the Two-Level Rider Choice Transit Network Design (2LRC-TND), which integrates machine learning and contextual stochastic optimization (CSO) to incorporate two levels of demand uncertainty.

Framework Details

The 2LRC-TND identifies travelers who regularly use public transit (core demand) and models the conditional adoption behavior of those who do not, based on the availability and quality of services. To capture these uncertainties, the system uses multiple travel mode choice models, based on machine learning algorithms. These models are integrated into a CSO, solved using a CP-SAT solver.

Validation and Results

The framework was evaluated in a case study in the Atlanta metropolitan area, involving over 6,600 travel arcs and more than 38,000 trips. The results demonstrate the effectiveness of the 2LRC-TND in designing transit networks that account for demand uncertainties and contextual information, offering a more realistic approach compared to traditional models.