請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/85432
標題: | 改善電池儲能系統溫度分布之進風口控制策略 Strategy for Improving Temperature Distribution of Battery Storage System through Inlet Condition Settings |
作者: | Yi-Wen Chen 陳怡妏 |
指導教授: | 楊鏡堂(Jing-Tang Yang) |
關鍵字: | 電池儲能系統,計算流體力學,熱管理,機器學習,分散式供氣, battery energy storage system,computational fluid dynamics,heat management,machine learning,distributed air supply, |
出版年 : | 2022 |
學位: | 碩士 |
摘要: | 本文分為三個部分,逐步建立適用於鋰離子電池儲能系統且具應用價值的分散式進口操控策略。第一部分參考資料中心的散熱方式,創新應用於電池儲能系統,具體改善現有商用儲能系統散熱效果不佳的熱流場因素。第二部分以電池模組的最高溫度、最大溫差作為指標,提出分散式的進口操控策略,有效消除局部熱點。第三部分提出兼具安全性與商業價值的散熱策略,引入機器學習演算法,驗證影響因素,並將電池的安全溫度、散熱系統的消耗功率納入控制策略的優劣評斷。 本文以商業軟體ANSYS Fluent進行數值模擬與分析,設計時引入資料中心提升散熱效率的方式且調整相異之處,比較三種進出風口配置,選定散熱效果最佳的配置作為本研究的物理模型,與本文參考的現有商用儲能貨櫃車之物理模型相比,最高溫度與最大溫差改善幅度達12.5 %和62.4 %。本研究解析各進口條件下的電池模組溫度,歸納出進口風量對於改善電池的最高溫度及最大溫差有最顯著效果,進口溫度僅對最高溫度改善較明顯。分析過程引入機器學習演算法協助印證兩進口條件對電池模組溫度分布的影響,建立各進口條件組合下,電池模組表面溫度是否超過安全溫度的預測。 更進一步以進口風量與進口溫度的影響為基礎,發展分散式進口控制策略,獨立設定進風口的風量或溫度。結果顯示,複合進口溫度控制策略可有效使目標機櫃的最高溫度降低8 °C以上,且未產生新熱點。與均勻調低進口溫度相比,此策略可省下10.99 kW的散熱系統消耗功率,為均勻調低進口溫度的37.1 %,大幅降低商業營運成本。 本文參考資料中心的散熱方式,改善商用的電池儲能系統配置;引入機器學習的演算法分析影響流場的因素,並建立電池儲能系統的分散式進口控制策略,有效節省空調系統與風扇的功率消耗,同時避免電池模組的溫度高於安全溫度。 This study proposes a control strategy based on the temperature distribution of battery modules and power consumption of air conditioning system, aims to improve the safety of battery energy storage system and profitability of air conditioning system. The first part of this study inspired by the cooling the solution for data center, based on the similarity between data center and battery energy storage system, this study re-designs the inlet and outlet of battery energy storage system. The second part establishes the distributed air supply strategy to avoid local hot spot, takes the maximum temperature and maximum temperature difference as the index for temperature distribution. The last part of the study duduces the strategy based on the power consumption, takes the profitability of air conditioning system into consideration. In this study I utilize ANSYS Fluent to calculate the flow field and temperature distribution. Inspired by data center, this study rearranges the configuration of battery energy storage system, the rearrangement shows the maximum temperature and the maximum temperature difference reduce 12.5% and 62.4%, respectively. The machine learning algorithm is utilized in the analysis of this study. Results show that the inlet volume flow rate has significant impact among the maximum temperature and the maximum temperature difference of battery modules. However, only the maximum temperature decreases as the inlet temperature decreases. To verify the influence of inlet conditions among the temperature distribution, the algorithm of decision tree is utilized in this study. The algorithm of support vector machine is used to establish a model, which can predict whether the surface temperature of the battery module exceeds the safe temperature under various inlet conditions. The distributed air supply strategy bases on the influence of inlet conditions is proposed in this study. The inlet volume and inlet temperature of each inlet will be assigned respectively. Results show that the compound inlet temperature control, which reduce the inlet temperature of two specific inlets, can at least cool down the target cabinet 8 °C without new hot spot. Compare to reduce all inlet temperature, the power consumption of the compound inlet temperature control is 13190.7W, 53.6% less then reduce all inlet temperature. This study rearranges the configuration of battery energy storage system with the solutions inspired by data center, and utilizes machine learning algorithm to decouple the influence factors of flow field. Moreover, this study builds up a distributed air supply strategy to maintain the module temperature within safety limit with less power consumption, reduces the cost of the ventilation system. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/85432 |
DOI: | 10.6342/NTU202201302 |
全文授權: | 同意授權(全球公開) |
電子全文公開日期: | 2023-08-01 |
顯示於系所單位: | 機械工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-0607202214424500.pdf | 31.12 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。