A Sequential Fusion Modelling Approach for Hypersonic Vehicles Based on Reinforcement Learning

Key Points

• Innovative sequential fusion modelling improves data synthesis efficiency for hypersonic vehicles, achieving cost reduction.
• Utilizing a Co-RBF multi-fidelity agent model significantly enhances model accuracy without needing excessive high-fidelity data.
• The method employs an active sampling strategy based on response surface extremes to optimize data collection and model predictions.
• This modeling approach demonstrates potential benefits for rapid iterations in hypersonic vehicle design and cost-saving opportunities.

Abstract