Improving Mathematical Modeling of Engineering Technologies for Smart Enterprises in a Machine Vision System

The article considers the issue of mathematical modeling of machine-building technologies of machine vision systems in the context of the transition to smart enterprises, which are the basis of modern Industry 4.0. It is noted that traditional approaches to modeling do not fully meet the requirements of digital production, which is characterized by high dynamism, the need for integration with cyber-physical systems, Internet of Things (IoT) technologies and adaptive control in real time. It is shown that modeling of machine-building technologies at a smart enterprise as a formalized approach is based on the specification of subsystems and a set of factors. The characteristics of mathematical models that can describe production processes at the enterprise with sufficient accuracy are given. In this case, the specifics of the structure, composition, key functional purposes and engineering, economic and its software and hardware aspects, as well as the conditions of machine-building production in accordance with the concept of a cyber-physical system, are taken into account. It is determined that according to the cyber-physical approach, the degree of automation of technological equipment varies from basic to high-level solutions. This requires modeling of production processes and technologies and the construction of mathematical models and their parameterization for further use in network interaction conditions when creating smart production and smart enterprises. It is found out that this requires the use of a machine (technical vision) system to ensure automation of product (service) quality control processes and management by processing and interpreting the information system. The possibilities of implementing a machine vision system at machine-building enterprises are identified. In this regard, such an artificial intelligence method as artificial neural networks is considered. The tasks in machine (technical) vision systems using the artificial neural network method are formulated. The advantages of this method in digital production and smart enterprises are determined. The integration of machine vision with robotic systems is clarified. A number of functions of the "machine vision-robot" complex and the possibilities of controlling the movement of robots during the task of controlling robotic platforms using a model of various types of artificial neural networks have been determined. The structure of an intelligent control system for robotic and mechatronic systems with a number of functional modules has been developed. An automated complex for the manufacture and assembly of parts and machines, intelligent production systems, have been considered at a smart enterprise. The functionality of intelligent production systems is based on modeling using mass service systems. Multiphase single- and multi-channel mass service systems are considered. The possibilities of models of automated production systems and the use of multi-level computer models in SCADA systems have been determined. The effectiveness of the created smart enterprises on the platform of a cyber-physical system and their integration into a single network of production systems of mechanical engineering have been shown. This involves the use of various automated design systems, technological preparation of production, a single database that will have a data management system about the manufacturer and mathematical modeling of advanced mechanical engineering technologies at smart enterprises.