This article is cited in 1 paper

Part 1. General problems of the intellectual systems theory

Learning algorithm of systems with discrete control

K. A. Golikov


Abstract: The learning algorithm was developed for the problem of positioning systems with discrete control, it is based on a method of generalizing using a global interpolation and a gradient descent of trials and fails that stored in the database. The algorithm is optimized by the criterion of reducing the learning time (number of attempts). The algorithm was tested on a simulator for models of systems operating on a plate of two different types: for a mobile differential-drive robot and for an open kinematic chain with rotational and prismatic joints.

Keywords: positioning, learning algorithm, robot, interpolation, approximation.