Abstract: An IoT-based energy information collection system for aluminum foil rolling workshops was proposed, taking into account the workshop′s process flow and energy consumption data characteristics. To address the issue of missing energy consumption data during the collection process, missing a data processing flow was designed, which utilized a WOA-CNN-LSTM hybrid neural network algorithm to interpolate the missing data. The system development employed technologies such as the MQTT protocol, JAVA language, and SpringBoot framework. It achieved energy information collection during the production process, information statistics, and supervision, improved data accuracy by addressing the issue of missing data during energy collection in the rolling workshop.

Key words: rolling workshop, energy information collection, missing data, WOA, CNN, LSTM