A Neural Network Learning System for Optimizing Selective
Laser Sintering Processes for Rapid Prototyping and Tooling Applications

Zhu Chun Bao and Wong David SK

German-Singapore Institute, Nanyang Polytechnic,
180 Ang Mo Kio Ave 8, Singapore 569830
Tel: (65) 550 1808, Fax: (65) 454 9871
Email: ZHU_Chun_Bao@nyp.gov.sg

FREE ABSTRACT (to read the full article, please log in below).

Selective Laser Sintering (SLS) is a new and one of the most important rapid prototyping techniques used in precision engineering industry. How to select optimal SLS parameters for building a desired SLS part is still a research and practical problem. The paper describes a development of a relational database (SLS-Database) and a neural network learning system (SLS-Learning System) for SLS parameter optimization. SLS-Database is used to store production data. The data stored not only provide production experience that will help the operator to set up proper parameters for prototyping applications, but also provide training data for the SLS-Learning System to learn the process. The trained neural networks can predict the output of the SLS prototyping application and based on the prediction, optimal selection of input parameters can be determined for building the desired SLS part. The paper also discusses effectiveness and further improvement of the SLS parameter optimization system.



To read the full article,
please log in:
If you haven't registered already,
you can do so for free:

© Copyright 2001 Professor F.R. Hall - University of Wolverhampton.