INTELLIGENT CONTROL SYSTEMS OF COBOT-TYPE MECHATRONIC SYSTEMS USING ML TECHNOLOGIES

Abstract

Goal. The works are research and development of the design of the manipulator hand, its mechatronic system and software control, with a special emphasis on the developed kinematics of the "Cobot" type bionic hand brush [1, 2]. Improvement of mechatronic systems control methods and development of control system algorithms for the creation of an innovative "Cobot" manipulator arm control system based on ML machine vision technologies and the Arduino platform.

Development of control algorithms in the Python programming language and their integration with Arduino using standard frameworks: "CV2", "MediaRipe", "Numpy" and "Serial" [3–6]. Coding of biomechanical joints for use in computer simulation modeling and remote control using artificial vision technologies.

Method. The research methodology included the use of mathematical methods to analyze the kinematics of the bionic hand and calculate the coordinates and kinematics of the mechanical system. The development of software control algorithms was based on the Python programming language. Standard frameworks: "CV2", "MediaRipe", "Numpy" and "Serial" are used to optimize the code. Software integration with an Arduino microcontroller was also used to control the mechatronic system.

The results. An innovative control system has been developed based on a "Cobot" hand manipulator, which responds to hand movements and gestures in front of the web camera. A block diagram of the interaction of electronic and electromechanical devices of the hand control unit of the "Cobot" hand manipulator was developed. Control algorithms have been implemented that can be applied in various fields of mechanical engineering, including medicine, industry, and the field of augmented and virtual reality. Algorithms for gesture recognition are presented, expanding the possibilities of using machine learning methods in various fields. The dependences of the kinematics of the fingers of the "Cobot" type bionic brush hand are given.

Practical significance. The obtained results indicate the powerful potential of using intelligent control systems of mechatronic devices. The system can be applied in medicine, industry and the field of augmented and virtual reality. The work contributes to the development of scientific platforms and is useful for further research in the field of mechatronics and intelligent control.