Tool based on artificial neural networks to obtain cooling capacity of hermetic compressors through tests performed in production lines
作者：
Highlights：
• An artificial neural network ensemble for estimating cooling capacity is proposed.
• Cooling capacity inference is made based on manufacturing data.
• Method based on bootstrapping and Monte Carlo is proposed for evaluation of uncertainty.
• The time required is about 0.05% of that using traditional measurement methods.
• Measurements are performed with reduced cost and known confidence interval.
摘要
•An artificial neural network ensemble for estimating cooling capacity is proposed.•Cooling capacity inference is made based on manufacturing data.•Method based on bootstrapping and Monte Carlo is proposed for evaluation of uncertainty.•The time required is about 0.05% of that using traditional measurement methods.•Measurements are performed with reduced cost and known confidence interval.
论文关键词：Artificial neural network ensembles,Hermetic compressors,Cooling capacity inference,Monte Carlo simulation,Bootstrapping
论文评审过程：Received 27 August 2020, Revised 6 December 2021, Accepted 31 December 2021, Available online 19 January 2022, Version of Record 24 January 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.116494
