Digital twin-based intelligent warehouse thermal environment simulation and automated control optimization

This study constructed a digital twin model of the intelligent warehouse thermal environment that integrates geometric, physical, and behavioral characteristics. Its accuracy was verified through multi-condition simulation experiments. The experiment involved 96 monitoring points deployed in a 50 m ? 30 m ? 12 m warehouse area over 72 hours. Results showed that the mean temperature simulation error was 0.21?C, with 87% of the monitoring points achieving an error of ? ?0.3?C. The highest accuracy was achieved around the equipment (mean 0.18?C). Humidity and temperature showed a significant negative correlation (correlation coefficient -0.76). Cardboard packaging absorbs 5% moisture by weight, increasing local humidity by 8% in closed aisles. The correlation is dynamic: -0.76 (dry conditions) vs. -0.5 (rainy, external humidity >90%). The weighting of the factors was equipment load (42%) > external temperature (35%) > cargo density (23%). The model accurately predicts thermal environment dynamics, providing support for efficient warehouse management and control.