An Intelligent Scheme to Emulate the Actions of Human Experts in DY Coil Winding Process

다국어 초록 (Multilingual Abstract)

The coil winding process in DY manufacturing process is known to be highly nonlinear, high-dimensional, and strongly coupled. The major concern of this paper is to build an intelligent scheme to emulate the human experts who adjust the winding parameters of the winding process to maintain the various convergence indices of DY within the prescribed range. To this end, we propose a hierarchical modular structure, in which a subsystem is designed for each convergence index and the adjustment action is synthesized by evaluating the outputs of the subsystems. The adjustment actions of the human experts are analyzed to build the adjustment rules for each subsystem based on fuzzy set theory and a modular neural network is used to evaluate the outputs of all the subsystem.

목차 (Table of Contents)

• Ⅰ. Introduction
• Ⅱ. DY Manufacturing Process
• Ⅲ. Moduar Structure for Emulating Human Experts
• Ⅳ. Fuzzy Rule Base for Local Decision Maker
• Ⅴ. Modeling of coil winding process using neural networks
• Ⅵ. Concluding Remarks
• References