Intelligent control system for an industrial manipulator

Fusaomi Nagata, Keigo Watanabe

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

In this chapter, two types of intelligent control systems are introduced. One is the adaptive learning technique for large-scale teaching signals. Recently various control methods as represented by PID control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. To overcome the problem, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of neural network by using back propagation algorithm, in general, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because of being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem concerning the learning performance, a simple and adaptive learning technique for largescale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator. The other is the fine gain tuning technique for model-based robotic servo controllers using genetic algorithms. Resolved acceleration control method and computed torque control method are used for nonlinear control of industrial manipulators, which are composed of a model base portion and a servo portion. The servo portion is a closed-loop system with respect to the position and velocity. On the other hand, the model base portion has the inertia term, gravity term and centrifugal/Coriolis term, which works for canceling the nonlinearity of manipulator. In order to realize high control stability, the position and velocity feedback gains used in the servo portion should be tuned suitably. In the latter half of the chapter, a simple but effective fine tuning method after the manual tuning process is introduced for the position and velocity feedback gains in the servo portion. At the first step, the search space for the gains are roughly narrowed down by a controller designer, e.g., considering the critically damped condition. At the next step, the gains are finely tuned within the space by using genetic algorithms. The genetic algorithms search for the better combination of the position and velocity gains. Simulations are conducted using the dynamic model of PUMA560 manipulator to validate the effectiveness of the proposed method.

Original languageEnglish
Title of host publicationIndustrial Control Systems
PublisherNova Science Publishers, Inc.
Pages117-151
Number of pages35
ISBN (Electronic)9781620816073
ISBN (Print)9781612099880
Publication statusPublished - Jan 1 2011

Fingerprint

Industrial manipulators
Intelligent control
Control systems
Controllers
Manipulators
Teaching
Tuning
Gravitation
Genetic algorithms
Dynamic models
Robotics
Acceleration control
Feedback
Backpropagation algorithms
Recurrent neural networks
Torque control
Three term control systems
Closed loop systems
Friction
Sampling

Keywords

  • Adaptive learning
  • Critically damped condition
  • Fine gain tuning
  • Genetic algorithms
  • Large-scale teaching signal
  • Model based servo controller
  • Neural networks
  • Nonlinear control
  • Puma560 manipulator
  • Servo system
  • Sigmoid function
  • Trajectory following control

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Nagata, F., & Watanabe, K. (2011). Intelligent control system for an industrial manipulator. In Industrial Control Systems (pp. 117-151). Nova Science Publishers, Inc..

Intelligent control system for an industrial manipulator. / Nagata, Fusaomi; Watanabe, Keigo.

Industrial Control Systems. Nova Science Publishers, Inc., 2011. p. 117-151.

Research output: Chapter in Book/Report/Conference proceedingChapter

Nagata, F & Watanabe, K 2011, Intelligent control system for an industrial manipulator. in Industrial Control Systems. Nova Science Publishers, Inc., pp. 117-151.
Nagata F, Watanabe K. Intelligent control system for an industrial manipulator. In Industrial Control Systems. Nova Science Publishers, Inc. 2011. p. 117-151
Nagata, Fusaomi ; Watanabe, Keigo. / Intelligent control system for an industrial manipulator. Industrial Control Systems. Nova Science Publishers, Inc., 2011. pp. 117-151
@inbook{8b998425ae634d3397b8577cd219353f,
title = "Intelligent control system for an industrial manipulator",
abstract = "In this chapter, two types of intelligent control systems are introduced. One is the adaptive learning technique for large-scale teaching signals. Recently various control methods as represented by PID control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. To overcome the problem, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of neural network by using back propagation algorithm, in general, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because of being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem concerning the learning performance, a simple and adaptive learning technique for largescale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator. The other is the fine gain tuning technique for model-based robotic servo controllers using genetic algorithms. Resolved acceleration control method and computed torque control method are used for nonlinear control of industrial manipulators, which are composed of a model base portion and a servo portion. The servo portion is a closed-loop system with respect to the position and velocity. On the other hand, the model base portion has the inertia term, gravity term and centrifugal/Coriolis term, which works for canceling the nonlinearity of manipulator. In order to realize high control stability, the position and velocity feedback gains used in the servo portion should be tuned suitably. In the latter half of the chapter, a simple but effective fine tuning method after the manual tuning process is introduced for the position and velocity feedback gains in the servo portion. At the first step, the search space for the gains are roughly narrowed down by a controller designer, e.g., considering the critically damped condition. At the next step, the gains are finely tuned within the space by using genetic algorithms. The genetic algorithms search for the better combination of the position and velocity gains. Simulations are conducted using the dynamic model of PUMA560 manipulator to validate the effectiveness of the proposed method.",
keywords = "Adaptive learning, Critically damped condition, Fine gain tuning, Genetic algorithms, Large-scale teaching signal, Model based servo controller, Neural networks, Nonlinear control, Puma560 manipulator, Servo system, Sigmoid function, Trajectory following control",
author = "Fusaomi Nagata and Keigo Watanabe",
year = "2011",
month = "1",
day = "1",
language = "English",
isbn = "9781612099880",
pages = "117--151",
booktitle = "Industrial Control Systems",
publisher = "Nova Science Publishers, Inc.",

}

TY - CHAP

T1 - Intelligent control system for an industrial manipulator

AU - Nagata, Fusaomi

AU - Watanabe, Keigo

PY - 2011/1/1

Y1 - 2011/1/1

N2 - In this chapter, two types of intelligent control systems are introduced. One is the adaptive learning technique for large-scale teaching signals. Recently various control methods as represented by PID control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. To overcome the problem, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of neural network by using back propagation algorithm, in general, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because of being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem concerning the learning performance, a simple and adaptive learning technique for largescale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator. The other is the fine gain tuning technique for model-based robotic servo controllers using genetic algorithms. Resolved acceleration control method and computed torque control method are used for nonlinear control of industrial manipulators, which are composed of a model base portion and a servo portion. The servo portion is a closed-loop system with respect to the position and velocity. On the other hand, the model base portion has the inertia term, gravity term and centrifugal/Coriolis term, which works for canceling the nonlinearity of manipulator. In order to realize high control stability, the position and velocity feedback gains used in the servo portion should be tuned suitably. In the latter half of the chapter, a simple but effective fine tuning method after the manual tuning process is introduced for the position and velocity feedback gains in the servo portion. At the first step, the search space for the gains are roughly narrowed down by a controller designer, e.g., considering the critically damped condition. At the next step, the gains are finely tuned within the space by using genetic algorithms. The genetic algorithms search for the better combination of the position and velocity gains. Simulations are conducted using the dynamic model of PUMA560 manipulator to validate the effectiveness of the proposed method.

AB - In this chapter, two types of intelligent control systems are introduced. One is the adaptive learning technique for large-scale teaching signals. Recently various control methods as represented by PID control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. To overcome the problem, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of neural network by using back propagation algorithm, in general, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because of being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem concerning the learning performance, a simple and adaptive learning technique for largescale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator. The other is the fine gain tuning technique for model-based robotic servo controllers using genetic algorithms. Resolved acceleration control method and computed torque control method are used for nonlinear control of industrial manipulators, which are composed of a model base portion and a servo portion. The servo portion is a closed-loop system with respect to the position and velocity. On the other hand, the model base portion has the inertia term, gravity term and centrifugal/Coriolis term, which works for canceling the nonlinearity of manipulator. In order to realize high control stability, the position and velocity feedback gains used in the servo portion should be tuned suitably. In the latter half of the chapter, a simple but effective fine tuning method after the manual tuning process is introduced for the position and velocity feedback gains in the servo portion. At the first step, the search space for the gains are roughly narrowed down by a controller designer, e.g., considering the critically damped condition. At the next step, the gains are finely tuned within the space by using genetic algorithms. The genetic algorithms search for the better combination of the position and velocity gains. Simulations are conducted using the dynamic model of PUMA560 manipulator to validate the effectiveness of the proposed method.

KW - Adaptive learning

KW - Critically damped condition

KW - Fine gain tuning

KW - Genetic algorithms

KW - Large-scale teaching signal

KW - Model based servo controller

KW - Neural networks

KW - Nonlinear control

KW - Puma560 manipulator

KW - Servo system

KW - Sigmoid function

KW - Trajectory following control

UR - http://www.scopus.com/inward/record.url?scp=85021917726&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021917726&partnerID=8YFLogxK

M3 - Chapter

AN - SCOPUS:85021917726

SN - 9781612099880

SP - 117

EP - 151

BT - Industrial Control Systems

PB - Nova Science Publishers, Inc.

ER -