REPRESENTATION AND LEARNING OF NONLINEAR COMPLIANCE USING NEURAL NETS

被引：32

作者：

ASADA, H

机构：

[1] Center for Information-Driven Mechanical Systems, Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge

来源：

IEEE TRANSACTIONS ON ROBOTICS AND AUTOMATION | 1993年 / 9卷 / 06期

基金：

美国国家科学基金会;

关键词：

Conformal mapping - Control nonlinearities - Control theory - Damping - Force measurement - Learning systems - Linear control systems - Neural networks - Nonlinear control systems - Robotics - Robots - Vectors;

D O I：

10.1109/70.265932

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

A new approach to compliant motion control using neural networks is presented. In the paper, ''compliance'' is treated as a nonlinear mapping from a measured force to a corrected motion. The nonlinear mapping by a multilayer neural network is represented, which allows us to deal with complex control strategies that cannot be represented by linear compliance, such as in stiffness and damping control.

引用

页码：863 / 867

页数：5

共 18 条

[1]

[Anonymous], 1987, LEARNING INTERNAL RE

[2] THE DYNAMIC ANALYSIS AND DESIGN OF A HIGH-SPEED INSERTION HAND USING THE GENERALIZED CENTROID AND VIRTUAL MASS [J].

ASADA, H ;

KAKUMOTO, Y .

JOURNAL OF DYNAMIC SYSTEMS MEASUREMENT AND CONTROL-TRANSACTIONS OF THE ASME, 1990, 112 (04) :646-652

[3]

ASADA H, 1989, 5TH P INT S ROB RES

[4]

ERNST HA, 1961, THESIS MIT

[5]

Hanafusa H., 1977, P IFAC S INFORMATION, P127

[6]

HIRZINGER G, 1983, 3RD P INT C ROB VIS, P549

[7] IMPEDANCE CONTROL - AN APPROACH TO MANIPULATION .1. THEORY [J].

HOGAN, N .

JOURNAL OF DYNAMIC SYSTEMS MEASUREMENT AND CONTROL-TRANSACTIONS OF THE ASME, 1985, 107 (01) :1-7

[8] A UNIFIED APPROACH FOR MOTION AND FORCE CONTROL OF ROBOT MANIPULATORS - THE OPERATIONAL SPACE FORMULATION [J].

KHATIB, O .

IEEE JOURNAL OF ROBOTICS AND AUTOMATION, 1987, 3 (01) :43-53

[9]

LOZANOPEREZ T, 1984, INT J ROBOTICS RES, V3

[10]

MASON MT, 1981, IEEE T SYST MAN MSC, V11

← 1 2 →