| URN | etd-0808124-161246 | Statistics | This thesis had been viewed 58 times. Download 1 times. |
| Author | Guan-Hao Peng | ||
| Author's Email Address | No Public. | ||
| Department | Institute Of Mechanical Engineering | ||
| Year | 2023 | Semester | 2 |
| Degree | Master | Type of Document | Master's Thesis |
| Language | zh-TW.Big5 Chinese | Page Count | 39 |
| Title | Using Big Data to Predict the Octane Number of Organic Compounds | ||
| Keyword | |||
| Abstract | When conducting chemical research, experimental operations are the best way to verify theories and predictions. However, experiments often require a lot of time and resources. Nowadays, AI (artificial intelligence) has become prevalent in daily life and can be widely used in various fields to mine data characteristics for constructing models. It might also be possible to use AI and big data to predict chemical reactions and molecular properties. By making predictions with big data before synthesizing new molecules, it is possible to determine whether the synthesis will be successful and avoid unnecessary expenses. Additionally, in industry, selecting the right catalyst is crucial for conducting reactions that are slow. Using big data to find the best catalyst could bring significant benefits. In gasoline, n-heptane tends to auto-ignite under high temperatures and pressures, causing knocking that reduces engine efficiency. N-heptane has an octane rating of 0, while iso-octane has an octane rating of 100. Other hydrocarbons have different octane ratings, which can be less than 0 or greater than 100. The octane rating of gasoline is directly determined by the composition ratio of various hydrocarbons in it. This research will utilize machine learning on data of the octane numbers of certain substances combined with big data to predict the octane numbers of chemical compounds. |
||
| Advisor Committee | |||
| Files | indicate access worldwide | ||
| Date of Defense | 2024-07-24 | Date of Submission | 2024-08-08 |