Enhanced LSTM framework for water-cooled chiller COP forecasting

摘要

The water-cooled chiller is one of the indispensable equipment in each building. Due to long-term use or improper operation, it is easy to cause components of the equipment to be wore out, generating additional energy consumption, and even causing malfunction. In this study, we used technology of IoT (Internet of Things) to upload the operating parameters and data in each period of every water-cooled chiller to the cloud, and presented the visualized result. Managers were able to understand the current status of each water-cooled chiller much easier with the visualized results, and at the same time we conducted predictive analysis on the collected water-cooled chiller data. If we can predict the COP (Coefficient Of Performance) value trend accurately, managers can effectively arrange personnel to carry out the maintenance work of each water-cooled chiller to reduce unnecessary energy and cost losses.