A multiple-input multiple-output (MIMO) fuzzy nets framework for part quality prediction and planning

Bufardi, Ahmed and Akten, Olcay and Arif, Muhammad and Xirouchakis, Paul and Perez, Roberto (2017) A multiple-input multiple-output (MIMO) fuzzy nets framework for part quality prediction and planning. WSEAS Transactions on Systems, 16. pp. 242-253. ISSN 2224-2678

[img]
Preview
Text (Bufardi-etal-TED2017-Towards-zero-defect-manufacturing-with-a-combined-online)
Bufardi_etal_TED2017_Towards_zero_defect_manufacturing_with_a_combined_online.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo

Download (1MB)| Preview

    Abstract

    In this paper, we present for the first time a general multiple-input multiple output (MIMO) fuzzy fuzzy-nets framework where the output parameters have different sets of inputs parameters and conflicting rules are generated. This is a significant departure from current fuzzy net approaches which are restricted to only one output. The proposed framework is based on a new model of the rule base architecture and of the predictive algorithm facilitating the resolution of conflicting rules. We demonstrate the application of the new framework for part quality prediction and input process parameter selection in Wire Electrical Discharge Machining (WEDM) with two part quality characteristics (surface roughness and average recast layer thickness) and three input process parameters (pulse-off time, feed rate and voltage); as a result a new technological paradigm for process planning in WEDM has been validated where not only surface roughness (the current situation) but also the average recast layer thickness is optimally planned.