Deep Reinforcement Learning Controls Rotating Detonation Engine Transitions

Researchers have developed a novel deep reinforcement learning (DRL) approach to control complex transitions in rotating detonation engines (RDEs). By reformulating the DRL problem within a moving reference frame that tracks the detonation wave, the system can effectively separate fast wave propagation dynamics from slower operating-mode changes. This method allows DRL controllers to more reliably induce rapid transitions between different engine states, outperforming controllers trained in a stationary frame and demonstrating broader effectiveness across various actuation periods and initial conditions. AI

IMPACT This research demonstrates a novel application of DRL for controlling complex multi-timescale physical systems, potentially advancing autonomous control in aerospace and other engineering fields.

RANK_REASON Academic paper detailing a new method for controlling complex systems using AI. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• deep reinforcement learning
• Kristian Holme

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →