With the acceleration of the digitization process and the continuous growth of data demands, traditional data centers often face issues of low energy efficiency, leading to increased energy consumption and energy costs. The cooling system is one of the main energy-consuming systems in data centers, therefore, energy-saving retrofit of the cooling systems has become a crucial topic, holding significant energy-saving potential. Modeling and optimizing the systems are particularly important for data center energy efficiency. This paper proposes a method to establish a comprehensive dataset for the entire energy-saving retrofit process, providing data support for device modeling.