請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/91294
標題: | 風力渦輪機維修排程與維修途程規劃之研究 Maintenance Planning and Routing Optimization for Wind Turbine Troubleshooting |
作者: | 洪智健 Derryadi Angputra |
指導教授: | 黃奎隆 Kwei-Long Huang |
關鍵字: | 風力渦輪機維修,服務團隊途程,最小化能源損失,排程演算法, Wind turbine maintenance,Service team routing,Minimize energy loss,Schedule algorithms, |
出版年 : | 2023 |
學位: | 碩士 |
摘要: | 本研究專注於開發針對台灣風力渦輪機的維護規劃和途程問題的解決方案,目標是減少可能發生的能源損失,我們稱之為風力渦輪機維護排程和途程問題(TMSRP)。風力渦輪機分佈在台灣各地的許多風場,這些風場風速變化不定且擁有不同數量的風力渦輪機。這些發電機可以根據其能源輸出劃分為三種類型, 1.500 kW, 2.000 kW, 和2.500 kW。如果發電機出現問題,可能會導致其立即停止運行,或者在特定到期日後如果未維修則停止運行。為了限制能源損失,一些擁有不同技能的團隊將負責排除故障,每個團隊都有預定的工作時間,從中央出發移動到風場之間,並考慮到旅行時間。研究的主要目標是規劃維護和任務途程,以最小化可能的能源損失,並考慮到維修處理時間、團隊技能和任務到期日。
本研究提出了一種基於潛在能源損失價值和團隊特性的排程演算法TMSRP-貪婪算法(TMSRP-GA),以產生日維修某排程和途程,並提出了兩種排程改善算法Weighted Randomness (WR) 和Iterative Max-Min Interchange (IMMI)。通過實驗結果,我們發現TMSRP-GA能夠構建一種相比於其他排程演算法有最小目標值的排程別外,WR有最低的計算時間和良好的整體降低目標函數。然而IMMI對初始解有依賴性,需要大量計算時間來收斂。 This research focused on developing a solution to the maintenance planning and routing problem for wind turbines in Taiwan, which aims to reduce the amount of energy loss that may occur which we called Turbine Maintenance Scheduling and Routing Problem (TMSRP). Wind turbines were built in numerous wind-farms spread across Taiwan, and these wind farms have different numbers of turbines with varying wind speeds that change unpredictably. The turbines can be categorized into three types based on their energy output. If a problem arises in a turbine, it could cause it to stop functioning immediately or after a particular due date if not repaired. To limit energy loss, teams with varying skills will handle the troubleshooting, each with a predetermined working hours, starting from a central warehouse and moving between wind-farms, accounting for travel time. The primary objective of the research is to plan maintenance and routing of tasks in such a way that minimizes potential energy loss by considering processing times, team skills, and task due dates. The study proposes a schedule construction algorithm TMSRP-Greedy Algorithm (TMSRP-GA) based on the potential energy loss worth and team characteristics to generate the schedule along with the routing for one day horizon, and also two schedule improvement algorithms WR and IMMI which then compared with baselines algorithm. Through the experimental result, we found out that TMSRP-GA was able to construct a schedule with a minimized objective value compared to other compared schedule construction algorithm, and WR also performed well and consistent with lowest computational time and good overall reduction in objective function. IMMI showed dependency to the initial solution and required a lot of computational time to converge. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/91294 |
DOI: | 10.6342/NTU202304138 |
全文授權: | 未授權 |
顯示於系所單位: | 工業工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-112-1.pdf 目前未授權公開取用 | 5.28 MB | Adobe PDF |
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。