Expert Formation of Functional Parameters Composition in Digital Mechatronic Systems
Autor: | V. M. Zaitsev, A. V. Gulay |
---|---|
Rok vydání: | 2019 |
Předmět: |
Sequence
Computer science Process (computing) Control engineering Regression analysis Collinearity Mechatronics Computer Science Applications Task (project management) Human-Computer Interaction Artificial Intelligence Control and Systems Engineering Electrical and Electronic Engineering Dimension (data warehouse) Reduction (mathematics) Software |
Zdroj: | Mekhatronika, Avtomatizatsiya, Upravlenie. 20:97-105 |
ISSN: | 2619-1253 1684-6427 |
Popis: | In the process of the mechatronic system functioning respective effective output parameters and parameters of its internal system state are generated. Preliminary selection of composition of the indicated parameters is a rather complicated task due to objective complexity of systems, availability of evident and latent interrelations between parameters. This task is complicated due to the necessity of simultaneous reduction of dimension and variation of control procedures and management in the mechatronic system. Solution of this problem may be attained with the aid of preliminary expert and empirical studies of complex models and mock-up specimens of designed mechatronic systems. In this regard, this work contains presentation of methodological aspects of adequately supported formation of composition of controlled parameters, their tolerance limits and other system features required for intellectual management. Rational sequence is described for performance of focused researches in order to reveal most important effective output parameters and parameters of internal system states. In accordance with the proposed algorithm the initial information and parameter picture of the intelligent mechatronic system is formed, studied and revised, its group expert evaluation and subsequent "thinning-out" of lists of its constituent parameters. A regression model is developed, which determines the form of interrelation between the parameters, as well as of the correlation model, which allows assessment of the volume of statistical interrelation between the system parameters. In the process of model analysis extremely weak dependences are detected between parameters, negative collinearity of factor variables is eliminated. |
Databáze: | OpenAIRE |
Externí odkaz: |