制冷技术2024，Vol.44Issue(5)：35-40,6.DOI:10.3969/j.issn.2095-4468.2024.05.106
基于和声搜索-长短记忆的NH3/CO2复叠制冷循环的最佳参数预测
Optimal Parameter Prediction of NH3/CO2 Cascade Refrigeration Cycle Based on Harmony Search with Long and Short Term Memory
摘要
Abstract
In order to study the relationship between the coefficient of performance(COP)of NH3/CO2 refrigeration cycle system and thermodynamic parameters,the high temperature condensation temperature and low temperature evaporation temperature of NH3/CO2 refrigeration cycle system is set,and the combination of Harmony search(HS)and long and short term memory(LSTM)neural network is adopted to find the best parameters in this paper.The LSTM is used to control the search accuracy of HS,and the convergence speed is fast.The heat transfer temperature difference between the optimal intermediate temperature and the optimal evaporative condenser at the maximum COP is obtained.When the condensation temperature of the high temperature stage remains unchanged at 30℃,while the evaporation temperature of the low temperature stage varies from-50℃to-25℃,the optimization result of the optimal intermediate temperature is-18.49℃to-9.15℃,and the heat transfer temperature difference of the optimal evaporation condenser is the lowest value in the optimization range.关键词
NH3/CO2复叠制冷/中间温度/极值寻优/和声搜索/长短记忆Key words
NH3/CO2 cascade refrigeration/Intermediate temperature/Extremal optimization/Harmony search/Long-short term memory分类
通用工业技术引用本文复制引用
吴兴应,蒯大秋,向夏楠,段姣姣..基于和声搜索-长短记忆的NH3/CO2复叠制冷循环的最佳参数预测[J].制冷技术,2024,44(5):35-40,6.基金项目
湖南省教育厅科学研究项目(No.22C0514) （No.22C0514）
湖南省自然科学省市联合基金(No.2021JJ50154). （No.2021JJ50154）
