| 摘要: | 2019年底至2020年初，全球爆發了嚴重特殊傳染性肺炎(COVID-19)疫情。為有效防止疫情蔓延，各國紛紛採取封城及封閉式管理措施，企業和學校不得不實行遠距離辦公和線上教學。從而催生"零接觸經濟"這一概念，然而， 2022年爆發的俄烏戰爭、美國頒布的"芯片法案"等事件迫使供應鏈面臨重組，使得供應鏈系統承受巨大壓力。因次，本文研究方法為針對該研究個案企業目前現行物料預測方法-由下往上法進行分析，並提出適合該公司的移動平均預測方法和平均絕對誤差(MAD)進行驗證，藉此協助該企業如何準確把握客戶和市場需求以降低庫存提出。針對系統層面提出先進規劃排程法進行研究，另一方面，透過該個案產業的歷史資料提出面對市場的不確定性如何能好的控管風險。本研究結果發現透過歷史資料分析該產業的景氣循環為每4年一次，同時透過移動平均法和平均絕對誤差驗證後，以2022年為例預測數量出機數2023台，實際出機數和為1843台兩者差距為9.8%，透過此計算方式得預測量出機數為2014台，兩者差距可縮小至9.2%。;At the end of 2019 and beginning of 2020, a severe outbreak of the novel coronavirus disease (COVID-19) occurred globally. To effectively prevent the spread of the epidemic, countries implemented citywide lockdowns and closed management measures, forcing companies and schools to adopt remote working and online teaching. This gave rise to the concept of a "zero-contact economy." However, events such as the 2022 Russia-Ukraine war and the U.S. CHIPS Act forced supply chains to reorganize, putting tremendous pressure on supply chain systems. Therefore, this study analyzes the current bottom-up material forecasting method used by the case study company, and proposes a suitable moving average forecasting method and mean absolute deviation (MAD) for verification to assist the company in accurately grasping customer and market demand to reduce inventory levels. At the system level, an advanced planning and scheduling method is proposed. On the other hand, historical data from the case study industry is used to propose how to better manage risks in the face of market uncertainty. The study finds that based on historical data analysis, the industry experiences a business cycle every 4 years. Using the moving average method and MAD verification, taking 2022 as an example, the predicted output quantity for 2023 is 2,023 units, while the actual output quantity is 1,843 units, a difference of 9.8%. Using this calculation method, the predicted output quantity is 2,014 units, narrowing the difference to 9.2%. |
