International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 03 | Mar 2024
p-ISSN: 2395-0072
www.irjet.net
Sensor Fusion Using Kalman Filter in Autonomous Vehicles Adarsh S. Puranik1, Swaroop Swamy K M2 ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - This research paper investigates the
The objectives of this research paper are twofold: first, to provide a comprehensive overview of sensor fusion using the Kalman filter in autonomous driving cars, including the underlying principles and methodologies; and second, to investigate the effectiveness of this approach through simulations and experiments in various driving scenarios. By achieving these objectives, this paper aims to contribute to the advancement of sensor fusion techniques in autonomous vehicles and pave the way for safer and more reliable autonomous driving systems.
application of sensor fusion using the Kalman filter in autonomous driving cars. The paper aims to explore the effectiveness of integrating data from multiple sensors to enhance the accuracy and reliability of state estimation in autonomous vehicles. The methodology employed involves a comprehensive review of the Kalman filter algorithm and its adaptation for sensor fusion in autonomous driving systems. The key findings reveal that sensor fusion using the Kalman filter significantly improves the accuracy of state estimation, leading to more robust autonomous driving capabilities. Moreover, the research highlights the importance of dynamic sensor fusion techniques in adapting to changing environmental conditions and improving the overall system reliability.
2. LITERATURE REVIEW Sensor fusion in autonomous vehicles has been a topic of extensive research, driven by the need for robust perception and decision-making capabilities in complex driving environments. Previous studies have explored various approaches to sensor fusion, aiming to integrate data from diverse sensor modalities to enhance the accuracy and reliability of perception systems in self-driving vehicles.
Key Words: Sensors, Sensor Fusion, Kalman Filter, Autonomous cars, Self-driving cars, Dynamic Sensor Fusion.
1. INTRODUCTION
One common approach to sensor fusion is the use of Bayesian filtering techniques, such as the Kalman filter, particle filter, and Bayesian networks. These techniques allow for the fusion of sensor data while accounting for uncertainties and noise, thereby improving the accuracy of state estimation. Additionally, machine learning-based methods, including neural networks and deep learning architectures, have been employed for sensor fusion tasks, leveraging large datasets to learn complex relationships between sensor inputs and vehicle states.
Autonomous driving, a rapidly advancing technology, holds the promise of revolutionizing transportation by providing safer, more efficient, and convenient mobility solutions. With the integration of artificial intelligence, advanced sensors, and computing systems, autonomous vehicles have made significant strides towards achieving full self-driving capabilities. However, one of the critical challenges in autonomous driving lies in accurately perceiving and understanding the surrounding environment to make informed driving decisions.
Several previous studies have specifically focused on the application of the Kalman filter in autonomous driving. For example, research by Thrun et al. (2006) demonstrated the use of the Kalman filter for sensor fusion in the DARPA Grand Challenge, where these vehicles are navigated through off-road terrain using a combination of GPS, inertial sensors, and laser range-finders. The study highlighted the effectiveness of the Kalman filter in integrating data from multiple sensors to accurately estimate vehicle position and orientation in challenging environments.
Sensor fusion, the process of combining data from multiple sensors, plays a pivotal role in addressing this challenge by providing a comprehensive and reliable representation of the vehicle's surroundings. By integrating information from sensors such as LiDAR, RADAR, Cameras, and GPS, sensor fusion enables them to perceive the environment in three dimensions, detect obstacles, track moving objects, and navigate safely and efficiently. The Kalman filter, a powerful recursive algorithm, serves as a cornerstone in sensor fusion for autonomous driving. Originally developed for aerospace applications, the Kalman filter has found widespread use in various fields, including robotics and autonomous vehicles, due to its ability to estimate the state of a dynamic system from noisy sensor measurements. In the context of autonomous driving, the Kalman filter facilitates the integration of sensor data to accurately estimate the vehicle's position, velocity, and orientation, thereby enabling precise navigation and control.
© 2024, IRJET
|
Impact Factor value: 8.226
Similarly, research by Ferguson et al. (2008) investigated the application of the Kalman filter for sensor fusion in urban driving scenarios, where these vehicles encountered complex traffic patterns and dynamic obstacles. By fusing data from LiDAR, radar, and vision sensors, the Kalman filter-based approach improved the vehicle's perception of its surroundings, enabling safe and reliable navigation in urban environments.
|
ISO 9001:2008 Certified Journal
|
Page 750