A dynamical systems approach to task-level system integration used to plan and control autonomous vehicle motion

https://doi.org/10.1016/0921-8890(92)90004-IGet rights and content

Abstract

Autonomous systems with multiple senory and effector modules face the problem of coordinating these components while fulfilling tasks such as moving towards a goal and avoiding sensed obstacles. We propose a set of organizational principles for dealing with this problem. The ideas are (a) to plan in terms of task-related variables that abstract from effector degrees of freedom and peripheral sensor coordinates but succinctly capture behavioral constraints; (b) to generate time courses of behavior through a dynamical system of the planning variables. Task constraints, such as targets to be reached, obstables to be avoided, etc. are expressed as parts of the planning dynamics in a principled fashion invoking concepts of the qualitative theory of dynamical systems. System integration is possible in the sense that all information provided by the various sensory modules and all information required by the various effector modules becomes part of the planning dynamics. Compression of such behavioral information is achieved in a second layer in which the relative strengths of different contributions to the planning dynamics are governed by competitive dynamics that separate convergent information, which is integrated by selecting a representative, from non-redundant information, which is kept invariant. The capability of the system to perform stable planning, make planning decisions, and integrate redundant as well as complementary information is demonstrated by software simulations. These include the simulation of control errors on both the effector and the sensor side.

References (30)

  • M. Eigen et al.

    The Hypercycle — A Principle of Natural Self-organization

    (1979)
  • T. Flash et al.

    The coordination of arm movements: An experimentally confirmed mathematical model

    Journal of Neuroscience

    (1985)
  • A.P. Georgopoulos

    Neurophysiology of reaching

  • J. Guckenheimer et al.

    Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields

    (1983)
  • Cited by (0)

    View full text