A Bottom-Up Approach to Socially Optimal Discrete Choices Under Congestion

Explore a 51-minute seminar on dynamic games and applications that delves into a bottom-up approach for constructing socially optimal discrete choices under congestion. Learn about a problem involving N agents with limited time to choose among finite destination alternatives, aiming to minimize collective energy expenditure while favoring motion strategies that limit crowding. Discover the three-stage solution process, including mapping optimal paths for arbitrary agent destination assignments, fixed fractions of agents, and identifying optimal fraction assignments. Examine the convex cost function as N approaches infinity, leading to simplified computations and epsilon-optimal decentralized control policies for large N. Gain insights from this joint work by Roland P. Malhamé, Noureddine Toumi, and Jérôme Le Ny, presented at the GERAD Research Center.