3rd Int’l Conf. on Recent Advances in Information Technology | RAIT-2016 | 978-1-4799-8579-1/16/$31.00 ©2016 IEEE Model of a Fuzzy Autonomation system for a Steel Wire Roll Mill L. N. Pattanaik Department of Production Engineering BIT, Mesra, Ranchi, India Email: lnpattanaik@bitmesra.ac.in Rajeev Agrawal Department of Production Engineering BIT, Mesra, Ranchi, India Email: rajeevagarwal@bitmesra.ac.in Lakshmi Kumari Department of Production Engineering BIT, Mesra, Ranchi, India Email: lakshmi19sept@gmail.com Abstract— In the present paper, an autonomation system based on fuzzy logic is proposed for a steel wire rolling mill to minimize a quality problem related to surface finish. Autonomation is a lean manufacturing concept that integrates automation with a human factor. The ability of self-diagnosis is achieved in the automated line by incorporating a computational intelligence tool like fuzzy logic. The decision support model proposed here provides intelligent decisive signals similar to the fuzzy capability of human brain. Fuzzy Logic control (FLC) model is designed to recognize the events that are likely to create defects and output action signals are generated. Speed difference between conveyor rollers and rolled products, percentage of carbon content and impact force on groove rollers are the three inputs and it produces two outputs in the form of either Andon (a visual signal) or line stoppage. By using Multiple Input and Multiple Output (MIMO) autonomation system, surface defects can be prevented by monitoring level of inputs. Keywords— fuzzy logic; autonomation; surface finish; MIMO system; Rolling mill I. INTRODUCTION Competitiveness in steel making industry is increasing due to the high demand. In order to protect the market share, automation of the manufacturing industrial process is essential and represents a challenge. Surface scratch is a major problem in automated line in steel manufacturing industry. For obtaining better quality steel and reducing waste, there is a need to adopt autonomation system. Several approaches have been proposed and being carried out by using different sensors and control system to avoid surface defects [1] but the system are not effective during their application. As referred through several works on defects of rolling mills, it can be found that making the system intelligent can result to achieve its purpose of avoiding surface scratches[2][3]. Design of such a logic based and intelligent controller demands the inclusion of fuzzy logic in the control system. Fuzzy logic control systems considered to serve the purpose because it is based on a logical system which is much closer to human thinking and action than traditional hard logical system. Fuzzy control systems describe the non-linear relationships between model parameters by mapping input variables to sets of membership functions. Fuzzy logic is similar to human thinking process and natural language enabling decisions to be made based on vague information. Fuzzy logic allows parameters to be described as having a certain membership degree in a set, while conducting fault diagnosis. A knowledge base, comprising rule base is built to support the fuzzy inference. Structure of fuzzy logic controller consists of following blocks: A. Inputs from sensors. Different inputs for FLC are taken through suitable sensors. B. Scaling Factor Scaling factor in a fuzzy logic controller is very important. Selection of suitable values for scaling factors are made based on the knowledge about the process to be controlled. C. Fuzzification Fuzzification is the conversion of crisp input from sensors to a degree of membership function. Linguistic terms are applied for each degree of membership that is applied to the input variables. D. Rule base and Inference Engine Set of rules made with the help of operator’s knowledge forms the rule base. Fuzzy inference is the system of designing the mapping from a given input to an output using fuzzy logic. The mapping then provides a platform from which decisions can be made. The process of fuzzy inference involves membership functions, fuzzy logic operators, and ‘IF X AND Y THEN Z’ rules. Fuzzy inference systems are of two types, that can be implemented in the Fuzzy Logic Toolbox. They are Mamdani-type and Sugeno-type. In this setup Mamdani-type fuzzy inference system is used. E. Defuzzification Defuzzification is defined as the conversion of fuzzy output to crisp output. There are many types of defuzzification methods available. Here we used Center of gravity (COG) method for defuzzification. Despite its complexity it is commonly used because, if the areas of two or more contributing rules overlap, the overlapping area is counted only once.