The Usage of Neural Networks for Motion Prediction of Autonomous Objects

Abstract

The problem of using neural networks to select the trajectory of an autonomous transport object as part of its control system is considered in this article. Environmental data from global and local sensors are used. A neural network is proposed to predict the movement of other road users to ensure the safety of the maneuver. A method for minimizing the functional characterizing the costs of movement is proposed. This allows the control system to work in true time.