Neuro-fuzzy based fusion of LiDAR and ultrasonic sensors to minimize error in range estimation for the navigation of mobile robots

摘要

Multi sensor data fusion plays a significant role in addressing the research problems for mobile robot navigation. In this paper, a fusion procedure was developed, based on an Adaptive Neuro Fuzzy Inference Systems (ANFIS), which fuses the data from ultrasonic and LiDAR sensors for a better range estimation as well as environment perception. The root mean square error analysis between the measured and actual distance across different experiment trials indicates that, ultrasonic sensor could provide the data with a root mean square error (RMSE) which varies between 0.3118 cm and 4.9953 cm, whereas the LiDAR sensor could provide the same between 5.1503 cm and 10.4773 cm over 0-4 m range. The RMSE of the proposed fusion algorithm varies between 3.4831 cm and 6.1471 cm. It can be observed that, the fusion process could reduce the mean square error present with the high cost LiDAR sensing system by one-half, while fusing it with the low cost ultrasonic sensing system. The fusion algorithm discussed in this paper provides a guidance to define various operating/safety zones for initiating necessary control action during the navigation of the mobile robots.