Page 142 - Vol.50
P. 142

廠務季刊  Facility Journal 2023


              制定標準守則 -MTU 發電機故障查修流程化
              MTU Generators Repairment Standardization

                                                                             文││何松穎 吳宸禾│台南廠務二部




                                                           摘要




                  摘要                                       於現今發電量綠能占比與日俱增、基載電力不足的環境
                                                           下,造就供電穩定度與天氣息息相關之結果,限電一詞因
                  1. 前言
                                                           而日漸頻繁出沒,各企業工廠急電穩定度亦因此日趨重
                   1.1  從發電機參數檢視機組狀態                       要。MTU 引擎於國內目前主力應用於海軍軍艦動力,其
                                                           電控式噴油系統低廢氣排放及參數監控能力強之優點遠近
                   1.2  狀態監測
                                                           馳名,反之其昂貴與壟斷性的保養耗材亦眾所皆知,如在
                   1.3  事後維修                               自我監測與查修階段即可辨識故障根本,除可提高系統運
                                                           轉穩定度,亦能同時達到節省維護成本之目的。本文旨在
                  2. 文獻探討
                                                           透過數據分析機組即時運轉資訊,達到即時預測故障與早
                   2.1  發電機運轉原理                            期修復之目的,維持發電機急電運轉可靠度。
                   2.2  分位數迴歸                              關鍵詞 : 發電機、急電、區間預測、大數據、故障分析、
                                                           分位數、線性迴歸
                  3. 研究方法
                   3.1  最佳化趨勢線                             Nowadays, under the circumstances of increasement
                                                           of renewable energy and the lack of based power,
                   3.2  百分位數迴歸                             power stability has been led to highly related with
                                                           weather.  Therefore,  the  term  power  ration  now
                  4. 結果與分析
                                                           appears frequently, and the stability of emergency
                   4.1  從油耗反映噴油嘴健康指標                       power  in  corporations  becomes  much  more
                                                           important. MTU engines are now mainly be used
                   4.2  應用分位迴歸於啟動油壓區間預測                    in navy�s ships. Its advantage of low emission is
                                                           famous for miles around. However, its expensive
                   4.3  應用分位迴歸於機組水溫區間預測
                                                           consumption  of  predictive  management  is  also
                  5. 結論                                    well-known. If users can identify engine fault at
                                                           first without helps, not only can system stability
                  參考文獻
                                                           be  upgraded,  but  also  the  purpose  of  saving
                                                           maintenance cost is achievable. This paper aims
                  作者介紹
                                                           to achieve two purposes. First, analyze and design
                                                           a database form for generators� parameters and
                                                           establish cross-phase platform for restoring fault
                                                           records. Second, create a reasonable interval of
                                                           prediction  for  users  to  monitor  timely  data  to
                                                           maintain the stability of emergency power.

                                                           Keywords : Generator, Emergency power, Interval
                                                           prediction, Big data, Fault analysis, Percentile,
                                                           Quantile, Linear regression













              ���
              ���
   137   138   139   140   141   142   143   144   145   146   147