Nonanalytical Methods for Motor Control

The following sections are included:

• PREFACE

• CONTENTS

The following sections are included:

• Analytical approach
  • Sets and mappings
  • Computer solutions
  • Ordered sets and metrics

• Algorithms and finite automata
  • Algorithms
  • Finite automata

• References

The following sections are included:

• Expert systems

• Neural networks

• Machine learning
  • Types of learning

• References

The following sections are included:

• Control task

• Solution of the control task

• Multilevel control

• Modeling the dynamics of human extremities

• References

The following sections are included:

• Modeling of muscles

• References

The following sections are included:

• Natural sensors

• The role of sensors in biological control

• References

The following sections are included:

• Biological basis for nonanalytical control

• Artificial reflex

• Organization of a skill-based expert system

• References

The following sections are included:

• Systems approach to the visual-motor coordination

• Reaching

• Tracking

• References

The following sections are included:

• Locomotion

• Finite state modeling of locomotion

• References

The readers may object to the prevailing attitude of the authors in insisting on the limitations of formal control theory when studying functional motions. Inadequate uses of analytical approaches and unjustified endowment of computers with anthropomorphic capabilities may serve as an excuse for such an attitude. Besides, the mathematical foundations of control theory are rigorously established and, hence, their scope clearly delimited. On the other hand, the non-analytical approach to the study of motor control which would rely on the representation of reflexes, motor skills, multilevel modeling of plants, specific optimization criteria, and connectionism with adaptivity and plasticity, is hardly outlined. Nonetheless, several important conclusions about the nature of motor control can be derived based on theoretical considerations and experimental results presented in the book…

The following sections are included:

• Methods

• Results

• Discussion

• References

The following sections are included:

• Model

• Results

• Discussion

• References

The following sections are included:

• Liapunov tracking concept (LTC)

• Results of the application of LTC

• References

The following sections are included:

• Functional electrical stimulation system for restoration of movements

• A finite state model of locomotion

• References

The work presented in this chapter follows the original idea on finite state control of locomotion [Tomović and McGhee, 1966], further developed by Tomović and collaborators Popović et al., 1991; Tomović et al., 1982; Tomović, 1984; Tomović et al., 1987; Turajlic and Drakulic, 1981]. A similar approach to control artificial legs can be found in the recent literature [Aeyels et al., 1992; Bar et al., 1983; James et al., 1991; Peeraer et al, 1991]. The most efficient controllers based on analytic methods for AKP follow the biomechanical considerations and respond to most of the requirements, but there have not been developed to the stage of clinical or home usage mostly due to the inappropriate technology used [Flowers et al., 1978; Koganezawa et al., 1987; Kuzhekin et al., 1984; Radcliffe, 1980; Stein and Flowers, 1980]…

The following sections are included:

• Principles of HASS operation

• Implementation

• References

At present most clinically applied functional electrical stimulation (FES) systems for restoration of locomotion use manually triggered open-loop control methods. These assistive systems integrate machine control which is triggered manually (volitionally), and a biological control where the subject corrects “errors” imposed by the imperfections of the machine control, and individual factors such as muscle fatigue, spasticity, joint contractures, muscle denervation, and environmental changes. Other FES systems use open-loop control based on the concept of “stored” sequences of muscle activations associated with the phases in the gait cycle. These sequences are based on the averaged EMG patterns recorded in able-bodied subjects while walking [Marsolais and Kobetič, 1983]. A direct, computer control of electrical stimulation was proposed [Chizeck, 1992] emphasizing muscular properties, but it is still not implemented in SCI subjects. Some investigators have introduced formal modelling to solve global issues such as standing stability, control strategies, feedback design [Khang and Zajac, 1989], and this approach is very promising…

The following sections are included:

• Model

• Methods

• Results

• Discussion

• References

Functional activation of paralyzed muscles by means of electrical stimulation of neuromuscular structures can restore limited grasping in humans who suffered spinal cord injury (SCI). There are two approaches which will be described in this chapter: 1) a method employing a set of implanted electrodes to activate individual muscles, the so called Case Western Reserve University (CWRU) grasping system [Peckham and Keith, 1992]; and 2) a somewhat less invasive method using surface electrodes for control of muscle groups…

The following sections are included:

• Methods

• Results

• References

The following sections are included:

• INDEX OF TERMS

• INDEX - AUTHORS