33th Congress of the International Council of the Aeronautical Sciences

03.4 - Applied Aerodynamics

MULTI-OBJECTIVE AERODYNAMIC OPTIMIZATION OF 2D HIGH-LIFT DEVICE BASED ON DISTRIBUTED DEEP REINFORCEMENT LEARNING

J.-H. Dai╣, P.-Q. Liu╣, L. Li╣; ╣Beihang University, China

We propose a method based on distributed reinforcement learning to solve the multi-objective optimization problem of high-lift devices. The new method draws lessons from Pareto ranking rather than merging multiple objectives into a single objective, so that a more uniform Pareto leading edge can be obtained. Combined with transfer learning, this method can improve the optimization efficiency.


View Paper