Abstract
Enclosed cabins play a vital role in aerospace, maritime navigation, and deep-sea exploration, where lighting critically impacts operator performance. This study explores the multimodal effects of lighting layouts on operators and develops an emotion prediction model based on electroencephalogram (EEG) features. Three lighting layouts (JL, XL, HL) and two illuminance levels (300 lux, 700 lux) were tested, recording operators' physiological signals, psychological states, and behavioral responses during cognitive tasks. The results show that the JL layout performs best in enhancing task performance and optimizing emotional states, especially under 700 lux conditions. The XL layout performs better under 300 lux but may increase physiological stress in high-illuminance environments. The HL layout enhances alertness but may increase cognitive load and induce negative emotions. Further analysis of EEG features and their combinations for emotion prediction revealed that event-related potential (ERP) features play a key role in emotion recognition, with either solo use or combined with other features improving classification performance. The k-nearest neighbor (k-NN) classifier showed the most stable performance in cross-validation and test accuracy, achieving the highest test accuracy of 87.5% with ERP features. This research provides theoretical and experimental support for optimizing lighting environments in enclosed cabins, improving personnel efficiency, and regulating emotions, offering valuable insights for task design and human-computer interaction optimization in confined environments.
| Original language | English |
|---|---|
| Article number | 113444 |
| Journal | Building and Environment |
| Volume | 284 |
| DOIs | |
| State | Published - 1 Oct 2025 |
Keywords
- Cognitive performance
- Emotion prediction
- Emotional intensity
- Enclosed cabin
- Lighting layout
- Physiological feature
Fingerprint
Dive into the research topics of 'Assessing lighting layouts in enclosed cabins: Multimodal impacts on operators and EEG-based emotion modeling'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver