Project title

Design a self driving car solution using Reinforcement Learning using Gym Environment

Submitted to:

  • Dr. Hirenkumar B Thakkar

Submitted By:

  1. Rushi Kanjaria1
  2. Kala Hemanth2
  3. K. Jhansi Naga Indusri3
  4. Apoorva Jindal4
  5. Nagasuri Alekhya5

Project Description

Nowadays, the technological Artificial intelligence trend is growing rapidly in the world. This paper can help one understand trends in autonomous vehicle technology and clear your thoughts on self-driving car. There are many different approaches to create a self-driving car, Deep Q-learning, Q-learning, and through machine learning also. However, our idea is very basic and simple when comparing with advance concepts. We preferred to create a simple small environment to make a simulation of our self-driving car model. The environment was provided by OpenAi gym which is a deterministic environment. OpenAi gym provided us the virtual simulation environment as well as the parameters needed to train and test the model. Wehave focused on two methods to test our model. The basic approach was to compare the performance of the car when tested using Q-Learning and when tested using Random Action, i.e., No rein-forcement learning. We have derived a theoretical model and analyzed on how to use Q-learning to teach car to drive. We have carried out simulation and on evaluating the performance, we found that Q-learning is a more optimal approach to solve the issue of self-driving car

Project Poster

Get Latest Notification about

Please ignore if you have already signed up.

Announcements, news and innovations!

From in your inbox.

By submitting this form, you are consenting to receive marketing emails from: Bennett University. You can revoke your consent to receive emails at any time by using the SafeUnsubscribe® link, found at the bottom of every email.