모터 동역학 기반 실시간 점성 마찰 추정 및 보상을 통한 보행 로봇의 적응 제어
Real-Time Viscous Friction Estimation and Compensation Based on Motor Dynamics for Adaptive Control of Legged Robots
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
It is a challenging topic for a legged robot to walk in various environments without changing the control parameters because of changes in viscosity depending on the environment. This paper presents an adaptive control technique for legged robots by estimating and compensating for viscous friction in each actuator using motor dynamics. The estimation leverages Recursive Least Squares (RLS) and the Gopinath method, providing a robust approach to determine viscous friction parameters in real time. To increase the estimation accuracy, the robot controller transmits the nonlinear terms generated in the robot to the motor drivers. Robot controller consists of Inverse dynamics-based controller and compensates for estimated viscous friction. To demonstrate the feasibility to robots and adaptability across diverse environments, the controller is tested on various walking speeds and in environments with different viscous friction, such as in water and air.
Abstract
It is a challenging topic for a legged robot to walk in various environments without changing the control parameters because of changes in viscosity depending on the environment. This paper presents an adaptive control technique for legged robots by estimating and compensating for viscous friction in each actuator using motor dynamics. The estimation leverages Recursive Least Squares (RLS) and the Gopinath method, providing a robust approach to determine viscous friction parameters in real time. To increase the estimation accuracy, the robot controller transmits the nonlinear terms generated in the robot to the motor drivers. Robot controller consists of Inverse dynamics-based controller and compensates for estimated viscous friction. To demonstrate the feasibility to robots and adaptability across diverse environments, the controller is tested on various walking speeds and in environments with different viscous friction, such as in water and air.