Moving Elements of Mobile Robots Stabilization Modelling

Abstract

The article describes the solution of the problem of stabilization of the mobile elements of mobile robots in transportation of goods. A neural network with different variants of perceptron by Python, tensorflow library and its extension tflearn was created. The simulation of the process of stabilization of the position of the mobile element of the technical system in the Open AI environment by means of the gym library is carried out. Data sets for neural network training are formed, training and testing of neural network by Python and gym library are carried out. As a result of testing recommendations on the choice of neural network architecture for the development of stabilization systems of mobile objects are given.