Octane number prediction for gasoline blends using convolution neural network / Zhu Yue

Zhu , Yue (2021) Octane number prediction for gasoline blends using convolution neural network / Zhu Yue. Masters thesis, Universiti Malaya.

[img] PDF ( The Candidate’s Agreement)
Restricted to Repository staff only

Download (655Kb)
    PDF (Thesis M.A)
    Download (1693Kb) | Preview


      With the development of information technology, the development of neural network plays an important role in the prediction of various situations in real life. At present, there are many prediction algorithms based on machine learning. According to the "80/20 rule" for building machine learning model, 80% of the time is spent of finding, cleaning, and organizing data, while the remaining 20% for training of the machine learning model. Machine performance learning models depend to a large extant to the data quality used train the model. Therefore, data preprocessing is widely considered to be one of the most critical stage in the whole process. In the project three commonly use algorithm are used for prediction of octane number for gasoline blends, which describes the behavior of the fuel in the engine at lower temperatures and speeds, and is an attemp to simulate acceleration behavior.These tree algorithm are back propagation (BP), radial basis funtion (RBF) and Extreme learning machine (ELM) algorithm. Simulation study on performance of these algorithm have been carried out with the available database. Simulation result show that RBF algorithm gives the best performance.

      Item Type: Thesis (Masters)
      Additional Information: Dissertation (M.A.) - Faculty of Engineering, Universiti Malaya, 2021.
      Uncontrolled Keywords: Machine learning; Prediction methods; Convolution neural network (CNN); Octane number prediction
      Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
      Divisions: Faculty of Engineering
      Depositing User: Mrs Rafidah Abu Othman
      Date Deposited: 13 Mar 2024 03:56
      Last Modified: 13 Mar 2024 03:56
      URI: http://studentsrepo.um.edu.my/id/eprint/14773

      Actions (For repository staff only : Login required)

      View Item