Indoor INS/LiDAR-Based Robot Localization With Improved Robustness Using Cascaded FIR Filter

Abstract

In this paper, an indoor inertial navigation system (INS) integrated with light detection and ranging (LiDAR) robot localization system is proposed to provide accurate information about the robot location. To achieve high accuracy and robustness, a cascaded finite-impulse response (FIR) filter is designed and incorporated into the proposed INS/LiDAR localization scheme. The cascaded scheme employs two FIR filters. An unbiased FIR filter is used to estimate the LiDAR-derived position by fusing distances between a robot and detected corner points. An extended FIR filter is used to fuse the LiDAR- and INS-based measurements. An experimental study indicates that the proposed scheme demonstrates higher robustness than the traditional methods of localization employing Kalman filtering.

Keywords

• Indoor localization
• inertial navigation system
• light detection and ranging
• FIR filtering
• Kalman filtering