Toward the use of bond graphs for manufacturing control: improving existing models

Autor(es)

Maluf, Arthur S. ; Sagawa, Juliana K. ; Borges, Roberto F. S. ; da Silva, Maira Martins

• Tipo de documento: Artigo Científico
• Fonte: Journal of the Brazilian Society of Mechanical Sciences and Engineering; v. 44, n. 11, p. 17-pg., 2022-11-01.

Resumo

In cyber-physical and automated manufacturing systems, parts scheduling and production control must incorporate feedback loops to consider the real-time shop floor state. Bond-graph modeling strategies allow modeling production control systems with feedback loops using flow and effort variables. Despite its applicability, there is a paucity of works applying this technique to this aim. The effort variable does not have a concrete interpretation in the few extant models. Often, buffers are modeled with infinite capacity, and a minimum function is used in the coupling interface between the machine and the buffer. These characteristics introduce nonlinearities, increasing the control design complexity. This paper proposes modifications to the existing bond-graph models for representing manufacturing entities to overcome these limitations, enhancing the usability of the modeling strategy. The proposal is applied to a 4-station shop floor as a proof of concept. The controllers are designed based on an optimal control approach coupled with parameters optimization so that engineers can adjust the capacity of the material sources and machines to desired levels. The proposal contributes to the literature by considering buffers with limited capacity, optimizing the control parameters, and designing responsive production systems with different shop floor configurations. These assets are relevant to Industry 4.0. (AU)

• Processo FAPESP: 19/12023-1 - Modelos dinâmicos para controle de sistemas de produção tipo job shop
• Beneficiário: Juliana Keiko Sagawa
• Modalidade de apoio: Auxílio à Pesquisa - Regular