The ice water system is composed of main equipment such as ice water host, cooling water tower and water pump. It is a huge and complex system. Each equipment interacts with each other through the circulation pipeline; and the same equipment also has different operating characteristics. Traditionally, thermodynamic formulas are often used, combined with traditional optimization search methods, to estimate parameters and find the best energy-saving operating point. However, the system operation mode established by this method is quite different from the actual complex ice-water system, and it is bound to be difficult to find the real optimal energy-saving point of the ice-water system. This study makes full use of the operation big data of the ice water system, introduces deep learning technology, and establishes the operation model of each subsystem of the ice water main engine, cooling water tower, and cooling water pump. Through the relationship between the input and output of each model, the optimal ice water host load rate, condenser inlet and outlet water temperature difference and cooling water tower cooling water outlet temperature are found, so that the overall ice water system can work at the lowest energy consumption point. And then achieve the purpose of energy saving. This article uses the actual operation data of the ice water system provided by Taiwan's AU Optronics panel factory as an example. The calculation results show that in terms of energy efficiency, compared with the total power consumption of the panel factory in 2019, the optimization of cooling water circulation parameters saved 0.89%. The ice water host load allocation optimization saves 0.33% of power consumption. In addition, the optimization of the total ice water temperature of the ice water host can help the panel factory's ice water system save 1.2% of power consumption in weather conditions where the dry bulb temperature is 25.8 and the wet bulb temperature is 19.4. This energy saving benefit will vary depending on the weather and the heat load of the factory area.
(Read the full text please subscribe to e-journals)
親愛的夥伴們~電子期刊第38期已經寄發囉！
Subscribers who have not completed the subscription on the publication date will be sent to your personal mailbox 3-5 days after the subscription.
★ If you do not receive the journal, please write to: tcta.info@gmail.com, we will deal with you as soon as possible, thank you!
