This outlines a method for dynamic pricing of high-occupancy toll lanes based on non-linear programming techniques, finite difference stochastic approximation, genetic algorithms and simulated annealing stochastic algorithms, coupled with vector autoregressive techniques, producing a solution for optimal flows and optimal tolls.

Don Graham is a Data Scientist and Consultant. He served as data science professor at Northwestern University and transportation professor at Florida Institute of Technology for over ten years, in addition to working in industry at The Institute for Defense Analysis, Lockheed, and AVIS.