광 변위센서를 사용한 선체 청소로봇의 위치추정 시스템 개발

Abstract

An industrial robot has been applied on the hull cleaning method to enhance an operational efficiency of entire cleaning process. Especially, autonomous robotic system is necessary for more efficient cleaning hull cleaning and the position estimation system is indispensible part in this system. Position estimation system of the hull cleaning robot, therefore, was studied for the autonomous hull cleaning process in this paper. Conventional position estimation method with rotary encoders is unsuitable for the hull cleaning robot on account of slippage between the robot wheel and the hull surface. Thus, a novel position estimation system using optical displacement sensors was suggested to solve this problem.

Operation environments and drive characteristics of the hull cleaning robot were analyzed to design the position estimation system effectively. Reflecting the results of the analysis, a position estimation algorithm which based on the dead reckoning and instantaneous center of rotation theory was developed. Performance test of the optical displacement sensor that measures the relative displacement with a contact-free optical sensor was implemented to find out the output characteristics according to the operating conditions including direction, speed, acceleration, height and surface type. In the position estimation system, two optical displacement sensors were used to reduce the measurement error and also data selection algorithm which choose more sensitive one in the two measured data was added to error reduction method. Furthermore, the monitoring PC operates the graphical based position estimation program that contains the position estimation algorithm. Consequently, the results of the position estimation are able to be displayed on the user interface screen in real-time and save on the database simultaneously.

The developed position estimation system was mounted on the scale model mobile robot which has an identical drive method with hull cleaning robot for experiments because the large scale support units, operation cost, high electric power, wide test area are required to operate the real hull cleaning robot. Experimental results demonstrate that the proposed position estimation system with the optical displacement sensors is more accurate compared to conventional system using rotary encoders.