The Department of Applied Mathematics weekly seminar is given by scholars and researchers working in applied mathematics, broadly interpreted.
Title: From diffusion models to stochastic control: a time-reversal methodology for feedback control design
Abstract: This talk introduces a novel methodology for constructing stochastic bridges in control-affine systems, effectively steering the system state from an initial condition to a desired target state within a finite time horizon. The proposed approach utilizes the time-reversal of diffusion processes to derive the required feedback control law. Subsequently, these stochastic bridges enable the design of feedback control laws that transition the system from an initial probability distribution to a specified target probability distribution, thereby addressing a Schrödinger-bridge-type problem. Preliminary numerical experiments on several benchmark scenarios are presented to demonstrate the applicability of the proposed approach.