Control of adaptive running platform based on machine vision technologies and neural networks

The paper considers the problem of selection of the optimal method for controlling an adaptive running platform for movement organization in virtual reality. The analysis of existing approaches for the control of such systems is carried out, a list of existing methods is formed, and new functions are developed. In order to solve the problem of user positioning within the framework of an adaptive running platform, two approaches, based on virtual reality trackers and using a machine vision, are considered and implemented. The problem of the study includes the choice of the optimal method of controlling an adaptive running platform to ensure maximum user comfort when walking on it. The study methodology includes the creation of a simplified model of human interaction with a running platform, the formalization of the assessment of the quality of movement, various methods of human positioning and platform control functions, followed by an experimental part on their comparison and the search for an optimal approach to running platform control. It was proved that machine vision technologies and neural networks allow positioning a person with sufficient accuracy, and they are deprived of the disadvantages of trackers. Moreover, comparative studies of six different control functions were carried out with the tracker-based positioning method and with the use of machine vision technologies. The most universal is the nonlinear one, a detailed zonal and proportional–differential functions are recommended for the tracker-based positioning method, linear and proportional–differential functions are recommended for positioning based on machine vision. The scientific novelty of the study consists in the formalization and comparison of various control methods of adaptive running platforms (based on linear and nonlinear functions, proportional differential law and neural networks), methods of positioning a person on them (using cameras and trackers), which will expand the area of knowledge about the optimal control functions of this class of devices. The practical significance of the research lies in a comprehensive description of the solution of the problems of organizing the control of adaptive running platforms and positioning a person by various methods.