Noise-Suppressing Newton Algorithm for Kinematic Control of Robots

Xiuchun Xiao,Dongyang Fu,Huan Wang,Jingwen Yan,Lin Wei

doi:10.1109/access.2019.2937686

Xiuchun Xiao, Dongyang Fu + Show 3 more

Open Access

https://doi.org/10.1109/access.2019.2937686

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2024
Citations: 22	License type: CC BY 4.0

Affiliation: Guangdong Ocean University, Shantou University

Abstract

In this paper, armed with the integral control method, a new noise-suppressing Newton (NSN) algorithm is proposed for the redundancy resolution of redundant robot manipulators efficiently. For practical hardware implementation, the discrete-time noise-suppressing Newton (abbreviated as DTNSN) algorithm is discretized from the continues NSN algorithm. Specifically, the distinguishing feature of the proposed DTNSN algorithm is that it can rigorously converge with inherent tolerance to noises induced by communication jamming and computational systematical errors. In contrast, considerable traditional algorithms often dispose of noises with the high-degree filter from the viewpoint of signal processing, which requires a complex system structure and further results in a heavy computational burden. Note that theoretical analyses are provided to elaborate the convergent property of the DTNSN algorithm polluted with constant bias, time-dependent linear noises and bounded random noises. Besides, by the proposed DTNSN algorithm, the end effector of both serial and parallel redundant robot manipulators complete the allocated motion planning and are impervious to the noisy simulated environment.

Highlights

T HERE exists a superior branch of robot arms referred to as redundant robot manipulators whose degrees of freedom (DOF) are more than that an end effector needs to conduct the allocated primary task [1]
Theorem 1: Under theoretically noiseless conditions, the steady-state residual error limk→∞ ε(tk) 2 synthesized by DTNSN algorithm (8) drops to O(l2), where vector O(l2) is made up of element O(l2) and · 2 denotes the Euclidean norm
The DTNSN algorithm (8) with the aid of the superior integral control method has been proposed for the manipulator redundancy resolution with the presence of unknown noises

Summary

INTRODUCTION

T HERE exists a superior branch of robot arms referred to as redundant robot manipulators whose degrees of freedom (DOF) are more than that an end effector needs to conduct the allocated primary task [1]. The NSN algorithm rapidly converges to the analytical solution to the redundancy resolution of redundant manipulators in the presence of noises. Extended solutions to the discrete-time NSN algorithm in serial and parallel redundant manipulators are provided in Section III with the comparable existing methods. Are provided for verifying the correctness and superiority of the proposed discrete-time NSN algorithm for redundant manipulators. 4) Synthesized by the newly proposed NSN algorithm, the experimental simulation of anti-noise manipulator motion planning is smoothly and precisely carried out in serial and parallel redundant manipulators

Redundant Manipulator Kinematics

NSN Algorithm Design

NSN ALGORITHM FOR REDUNDANCY RESOLUTION

THEORETICAL ANALYSES AND RESULTS

SIMULATIVE EXPERIMENTS

CONCLUSIONS