Share

Application of a support vector machine algorithm for improving effects of defrosting of commercial refrigerated display cabinets in supermarkets.

Author(s) : CAO Z. K., ZHANG C. X., GU B.

Type of article: Article

Summary

This article presents a prediction method for improving effects of defrosting of open vertical refrigerated display cabinets in supermarkets that is based on a support vector machine (SVM) algorithm. To determine the refrigerating period (RP) and the defrost off period (DP) corresponding to different indoor climate classes, models for total energy consumption/total display area (TEC/TDA) and average mean temperatures of all M-packages (TM), associated with the defrosting effects of display cabinets, are proposed. After the training and validation data sets were constructed from modified Box and Behnken design experiments, the constructed object functions were solved using a SVM algorithm with different input parameter combinations. As a result, within 1°C for the variation of t DP , the TEC/TDA and the mass of condensate obtained after defrosting, achieved from the experiments by using the predicted combination of the controlled parameters, were found to be reduced by 27.0%/15.6% and 27.2%/15.4% under 3M/0M limate class, respectively.

Details

  • Original title: Application of a support vector machine algorithm for improving effects of defrosting of commercial refrigerated display cabinets in supermarkets.
  • Record ID : 30008845
  • Languages: English
  • Source: HVAC&R Research - vol. 19 - n. 3
  • Publication date: 2013/04
  • DOI: http://dx.doi.org/10.1080/10789669.2012.758534

Links


See other articles in this issue (6)
See the source

Indexing