二氧化碳過度排放造成地球暖化，造成我們所處的環境岌岌可危，而二氧化碳的排放主要來自於化石燃料的產生，其中大部分為提供發電所用，因此節能省電對我們的環境來說是非常重要的議題。訂定契約容量是電力公司能穩定供給能源與規劃未來新電廠建設的依據，電廠評估至建設時間至少需五至十年，若是能降低契約容量，則可減少電力公司建設新電廠的需求，進而達到減少發電量的產生，並有效降低二氧化碳的排放量。 過去有許多關於契約容量的研究，其中可分為兩個面向，一個為供電端的智慧電網，另一方面則為用電端的研究。供電端過去的研究主要探討電力公司在區域性應施行何種政策才能達到降低契約容量與節能省電的目標，如需量競價等，但是這些政策若沒有大多數大用電量的企業配合，則難以達到降低契約容量的目的。而用電端主要有三種研究，第一種為根據過去的用電資料，分析企業應與電力公司簽訂多少契約容量才能達到最小化電費的效果，此類方法並無法降低契約容量；第二種為降低閒置設備，以增加用電效率，同樣無法達到減少契約容量；第三種則透過視覺化用電情況，使生產者隨時注意是否超過最大需量，並適時延後或提早生產，此方法只要生產者一個閃失，或是臨時需要加速生產大量用電，根據契約容量之規定，則會非常輕易的超約並產生罰款。 因此本研究利用原料與用電相關性高的特性，提出透過交錯生產穩定排程原料的控制並以 DBSCAN 分群法將生產特性相近的品項合併生產，以達到穩定用電的需求，且提供生產者簡易的方式根據排程進行生產，藉此達到實質降低契約容量之效果。最後，將此方法應用至台灣某汽車零組件主要供應商之資料，預估至少節省契約容量1000KW，並預估達到一年減少 5,431 噸二氧化碳排放量。;The global warming caused by excessive carbon dioxide emissions has severely damaged our environment. Carbon dioxide emissions mainly come from the burning of fossil fuels, most of which are used for providing thermal power. Therefore, reducing power generation is vital for environmental sustainability. To conserve power generation, power companies encourage companies to set up contract capacity to cap the maximum consumption of powers. If companies can reduce contract capacity, the demand for power generation can be reduce and so does carbon dioxide emissions. The studies of contract capacity can be divided into two aspects. One is the design of smart grid and the other is the reduction of power consumption. The smart grid related research discussed the implementation of the mechanism to share electricity. The past research on the power consumption can be further divided into three areas. The first is to average the existing power consumption to set up the contract capacity. The second is to fully utilize the contracted capacity by increasing the utilization rate of equipment. The third is to visualize the electricity consumption to help users control the usage of electricity. However, none of the above mentioned approaches tries to minimize the contracted capacity by averaging the amount of working orders. This study proposes to stabilizes the control of raw materials in the scheduling through staggered production methods, and using the DBSCAN method to combine production items with similar production characteristics to achieve stable electricity demand. Then, it can provide producers with an easy way to produce by schedules. A case study is being conducted on a major automotive component supplier in Taiwan. the result shows that the contract capacity can save 1000 KW per year and reduce 5,431 tons of CO2 emissions as a result.
