Share

IIR document

Machine learning model of regenerative evaporative cooler for performance prediction based on experimental investigation.

Author(s) : GUPTA A. K., KASHYAP S., SARKAR J.

Type of article: IJR article

Summary

The regenerative evaporative cooler is experimentally investigated in this study and the predictive machine learning model is developed based on the test data to predict the energy and exergy performances of the device. This statistical machine learning framework is used to study the effect of six input operating variables (air inlet temperature, inlet air flow rate, air inlet specific humidity, water flow rate, inlet water temperature and extraction ratio) on the output variables (supply air temperature, cooling capacity, dew point effectiveness, coefficient of performance and exergy efficiency). The linear and polynomial regression models are followed to see the exactness of the model and check the train score against the test score. The effect of all the input variables on the output variables are discussed as well. In the case study, the developed machine learning model is used to predict the performances of fabricated regenerative evaporative cooler based on the field weather data of different days of 2020 in Varanasi. Model train score is 0.9974; the test score is 0.9912, the mean squared error is 0.1427, and the root mean squared error is 0.3778. This model can be used to predict the cooler performance under varied operating conditions.

Available documents

Format PDF

Pages: 178-187

Available

  • Public price

    20 €

  • Member price*

    Free

* Best rate depending on membership category (see the detailed benefits of individual and corporate memberships).

Details

  • Original title: Machine learning model of regenerative evaporative cooler for performance prediction based on experimental investigation.
  • Record ID : 30029538
  • Languages: English
  • Source: International Journal of Refrigeration - Revue Internationale du Froid - vol. 137
  • Publication date: 2022/05
  • DOI: http://dx.doi.org/10.1016/j.ijrefrig.2022.02.006
  • Document available for consultation in the library of the IIR headquarters only.

Links


See other articles in this issue (23)
See the source

Indexing