Loading...

Active Control of Drag Force in Automobile in Order to Reduce Fuel Consumption

Mehrafsar, Moein | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57323 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Boroushaki, Mehrdad
  7. Abstract:
  8. In this thesis, an attempt has been made to provide a smart and innovative system to optimize fuel consumption in cars by using the combination of engineering sciences, aerospace, and machine learning. Considering the current challenges of societies, the economic importance of saving fuel, as well as recent developments in the fields of machine learning and reinforcement learning, this research tries to improve the performance of cars in order to minimize the air resistance force of car bodies. In this regard, deep neural networks have been used to learn the dynamics of aerodynamic forces and reinforcement learning algorithms, especially DDPG (Deep Deterministic Policy Gradient), have been used to optimize control angles to minimize fuel consumption. This system has been able to not only accurately model the drag force on the car, but also optimize the air resistance force by adjusting the control parameters. In the first stage, with a complete review of the research literature in the field of artificial intelligence and aerodynamics, the differences and integration of these two fields have been examined and the strengths of each have been used. In the following, the steps of the project implementation including CFD (Computational Fluid Dynamics) modeling for airflow analysis, development of deep neural networks as a prediction model of aerodynamic forces, and optimization of control angles using the DDPG algorithm are explained in detail. The results obtained from the present research show a significant improvement in the efficiency of the proposed system. This thesis has been completed with a critical study and critique of the results, suggestions for future research and practical suggestions to increase the efficiency and applicability of the proposed system
  9. Keywords:
  10. Fuel Consumption ; Machine Learning ; Deep Neural Networks ; Reinforcement Learning ; Aerodynamic Power ; Fuel Consumption Optimization

 Digital Object List

 Bookmark