No abstract

No abstract

This work is devoted to the prediction of human actions. The article proposes a machine learning algorithm for predicting human physical actions, created using a software package for collecting data, processing data and training a regression algorithm on the processed data. The results obtained are associated with the automatic determination of the beginning and type of physical action performed by a person. The work is aimed at improving control systems for industrial use of exoskeletons designed to increase human strength through an external frame. In the future, it is possible to use the research results for better interaction with assistive devices in enterprises.

This work describes development of a movement simulation system for autonomous car "Niva" with a detachable self-driving system on a multi-lane road. Models for various obstacles and shrinkings of road are presented, as well as methods for more flexible description of driver behavoir on the road. Results of numerical experiments are presented..