Principles of Building an Intelligent System for Structural Synthesis of Mechanical Engineering Objects

Abstract

The article considers an original classification of automation systems for structural synthesis (design) ofmechanical engineering objects, allowing not only to identify (classify) existing automation systems for design, but also to determine the direction of development of structural synthesis systems, to substantiate fundamentally new approaches to create such systems, and to show the possibility of creating intelligent tool environments for building automated design systems for complex engineering objects, whose basic operation algorithm will be based not on functional subject area dependencies, but on optimized search algorithms. Preparation of information support and system training process to follow the basic principles of modularity (describing the synthesis object as a set of discrete structures, elements, and characteristics), decomposition (a basic tool for forming a correct set of modules), hierarchical representation (the need to divide an object into substructures in a step-by-step manner during the formation of a set of modules, with the formation of a set of structure elements proceeding from the top down, from the product to the nodes, sub-nodes, assemblies, sub-assemblies, and parts, while the characteristics are formed from the bottom up), emergence (the correct reference of characteristics to structure elements), sufficiency (determining the depth of source material analysis during decomposition), and consistency (recording possible causes of inconsistency and using them correctly), is an essential condition to implement the approach. The article provides a detailed description of these principles, the need to comply with them, and their impact on the information quality provided during the system's training. It also highlights the importance of using these principles in a comprehensive manner, their correlation and applicationfeatures.