Abstract
This study proposes an optimization model for museum exhibition space design based on Artificial Intelligence (AI) technology, aimed at enhancing the flow and interactivity of museum visits. The model integrates reinforcement learning, computer vision (CV), and affective computing to optimize spatial layout and interactive design, thereby improving both visit efficiency and audience experience. Experimental results show that the optimized layout increases spatial fluency by 18.1%, reduces congestion, and improves visitor efficiency. The exhibit visit rate rises by 50.0%, indicating that the design more effectively draws attention to a wider range of displays. Additionally, affective computing enables real-time emotion recognition and adaptive feedback based on data such as facial expressions, vocal tones, and body posture, further enhancing personalization and interactivity. Overall, the AI-based model significantly outperforms traditional design approaches in improving the visitor experience. This study offers a novel approach to intelligent museum design and outlines promising directions for future research.