역분산 가중평균을 이용한 다중 관성 센서 기반 관로 탐사기법 개발

Abstract

We propose a method for developing an in-pipe inspection robot based on multiple inertial sensors. Estimating the position of underground pipelines where satellite signals do not reach remains challenging. High-precision inertial sensors and high-tech mobile robots can be solutions, but their high price limits their general use. We developed an in-pipe inspection robot by combining various low-cost sensors with a microcomputer-based RC car platform. First, we fabricated a multi-inertial sensors module by combining commercial grade low-cost MEMS inertial sensors. The sensor values measured by the multi-inertial sensor are transmitted to the main computer through the MCU, and the attitude angle of the vehicle is finally calculated through the inverse variance weighted average. The travel distance of the robot is estimated by using hall sensors and neodymium magnets attached to the inside of the wheels. Also, we measured the pipe diameter using multiple ultrasonic sensors. We verified the estimation accuracy of each sensor through experiments and consequently estimated the 3D trajectory of the in-pipe robot.