The effect of artificial neural network towards the number of particles of Rao-Blackwellized particle filter using laser distance sensor

Jamaludin, Amirul and Norhidayah, Mohamad Yatim and Mohd Noh, Zarina (2023) The effect of artificial neural network towards the number of particles of Rao-Blackwellized particle filter using laser distance sensor. International Journal of Advanced Computer Science and Applications, 14 (1). pp. 691-700. ISSN 2158-107X

Text 0127823022023.PDF Download (1MB)

Abstract

Rao-Blackwellized particle filter (RBPF) algorithm aims to solve the Simultaneous Localization and Mapping (SLAM) problem. The performance of RBPF is based on the number of particles. The higher the number of particles, the better the performance of RBPF. However, higher number of particles required high memory and computational cost. Nevertheless, the number of particles can be reduced by using high-end sensor. By using high-end sensor, high performance of RBPF can be achieved and reduced the number of particles. But the development of the robot came at a high cost. A robot can be equipped with low-cost sensor in order to reduce the overall cost of the robot. However, low-cost sensor presented challenges of creating good map accuracy due to the low accuracy of the sensor measurement. For that reason, RBPF is integrated with artificial neural network (ANN) to interpret noisy sensor measurements and achieved better accuracy in SLAM. In this paper, RBPF integrated with ANN is experimented by using Turtlebot3 in real-world experiment. The experiment is evaluated by comparing the resulting maps estimated by RBPF with ANN and RBPF without ANN. The results show that RBPF with ANN has increased the performance of SLAM by 25.17% and achieved 10 out of 10 trials of closed loop map by using only 30 particles compared to RBPF without ANN that needs 400 particles to achieve closed loop map. In conclusion, it shows that, SLAM performance can be improved by integrating RBPF algorithm with ANN and reduces the number of particles

Item Type:	Article
Uncontrolled Keywords:	SLAM, occupancy grid map, Rao-Blackwellized particle filter, artificial neural network, laser distance sensor
Divisions:	Faculty of Electronics and Computer Engineering
Depositing User:	Norfaradilla Idayu Ab. Ghafar
Date Deposited:	19 Jun 2024 09:38
Last Modified:	19 Jun 2024 09:38
URI:	http://eprints.utem.edu.my/id/eprint/27096
Statistic Details:	View Download Statistic

Actions (login required)

View Item