Cavendish Series: Data Challenges and Physics-Informed Time-Series Reinforcement Learning for Adaptive Optimisation of Steelmaking Processes
You are invited to join us for the 14th webinar in the Cavendish Lecture Series, where the focus will be Data Challenges and Physics-Informed Time-Series Reinforcement Learning for Adaptive Optimisation of Steelmaking Processes
You will hear from
Professor Hongbiao Dong FREng, FIMMM
What to expect
Steelmaking is a dynamic, nonlinear and partially observable process in which operational decisions must respond continuously to changing conditions. Industrial data are often heterogeneous, asynchronous, noisy and incomplete, and are affected by variable feedstocks, sensor drift and changing operating regimes. This talk presents a physics-informed time-series reinforcement learning framework for addressing these data challenges while enabling adaptive process optimisation.
Previous webinars in this series
