Abstract

Machining is always accompanied by many difficulties like tool wear, tool breakage, improper machining conditions, non-uniform workpiece properties and some other irregularities, which are some of major barriers to highly-automated operations. Effective tool condition monitoring (TCM) system provides a best solution to monitor those irregular machining processes and suggest operators to take appropriate actions. Even though a wide variety of monitoring techniques have been developed for the online detection of tool condition, it remains an unsolved problem to look for a reliable, simple and cheap solution. This research work mainly focuses on developing a real-time tool condition monitoring model to detect the tool condition, part quality in machining process by using machine learning techniques through sound monitoring.

The present study shows the development of a process model capable of on-line process monitoring utilizing machine learning techniques to analyze the sound signals collected during machining and train the proposed system to predict the cutting phenomenon during machining. A decision-making system based on the machine learning technique involving Support Vector Machine approach is developed. The developed system is trained with pre-processed data and tested, and the system showed a significant prediction accuracy in different applications which proves to be an effective model in applying to machining process as an on-line process monitoring system. In addition, this system also proves to be effective, cheap, compact and sensory position invariant. The successful development of the proposed TCM system can provide a practical tool to reduce downtime for tool changes and minimize the amount of scrap in metal cutting industry.

Library of Congress Subject Headings

Machining--Automation; Machine learning

Publication Date

7-19-2017

Document Type

Thesis

Student Type

Graduate

Degree Name

Mechanical Engineering (MS)

Department, Program, or Center

Mechanical Engineering (KGCOE)

Advisor

Rui Liu

Advisor/Committee Member

Patricia Iglesias

Advisor/Committee Member

Raymond Ptucha

Campus

RIT – Main Campus

Plan Codes

MECE-MS

Download

Share

COinS