Double Deep Q-Learning and Faster R-CNN-Based Autonomous Vehicle Navigation and Obstacle Avoidance in Dynamic Environment.

Razin Bin Issa,Md Golam Rabiul Alam,Modhumonty Das,Kazi Shah Nawaz Ripon,Md Khalilur Rhaman,Monika Barua,Md Saferi Rahman

doi:10.3390/s21041468

Razin Bin Issa, Md Golam Rabiul Alam + Show 5 more

Open Access

https://doi.org/10.3390/s21041468

Copy DOI

Abstract

Autonomous vehicle navigation in an unknown dynamic environment is crucial for both supervised- and Reinforcement Learning-based autonomous maneuvering. The cooperative fusion of these two learning approaches has the potential to be an effective mechanism to tackle indefinite environmental dynamics. Most of the state-of-the-art autonomous vehicle navigation systems are trained on a specific mapped model with familiar environmental dynamics. However, this research focuses on the cooperative fusion of supervised and Reinforcement Learning technologies for autonomous navigation of land vehicles in a dynamic and unknown environment. The Faster R-CNN, a supervised learning approach, identifies the ambient environmental obstacles for untroubled maneuver of the autonomous vehicle. Whereas, the training policies of Double Deep Q-Learning, a Reinforcement Learning approach, enable the autonomous agent to learn effective navigation decisions form the dynamic environment. The proposed model is primarily tested in a gaming environment similar to the real-world. It exhibits the overall efficiency and effectiveness in the maneuver of autonomous land vehicles.

Highlights

Autonomous vehicles, in this modern era, are a vital part of an advanced transportation system
The integration of Faster R-Convolutional Neural Network (CNN) and Double Deep Q-Learning Network (DDQN) shows the optimum result in autonomous exploration
The data-set that is used here to train the image classifier shows the accuracy of 94.06%. This accuracy level represents the effectiveness of the classifier to identify any object which may appear before the autonomous vehicle

Summary

Introduction

Autonomous vehicles, in this modern era, are a vital part of an advanced transportation system. Autonomous vehicles are considered to be one of the fastest-growing technologies that exist at present. The autonomous vehicle extracts environment perception to conclude directing the agent [1]. Decision-making is the main module of an autonomous vehicle. It is vital to make an autonomous vehicle learn finding an optimal path for traversing. This work suggests the integration of Reinforcement Learning method in an autonomous vehicle to make it able to take optimal decisions while traversing in a dynamic environment

Results

Discussion

Conclusion