Walking robot: a non-deterministic model of control

Abstract

The development of a model of motion control for walking robots is reported. The goal of the project is the elaboration of methods to construct the tree main levels of a control system of the robot's motion and a program complex to realise the robot's behaviour simulation in complex topography. A notable amount of achievements has been reached by now in developing the model of behaviour control for robots in various activity areas. Most of the models, however, are, to our mind, too deterministic. The hope to obtain interesting results in the field with a model based, in a sense, on the principle of maximum indetenninism gave rise to the project 'discussed in the present paper. The corresponding ideology has been almost literally transferred here from the asynchronous programming theory; as to the walking, it was chosen as a model domain because it offers combination of both general and special problems.