FreeTechBooks.com Homepage
FreeTechBooks.com
Free Online Computer Science and Programming Books, Textbooks, and Lecture Notes


Probabilistic Design
Reply with quote
Probabilistic Design

Author : John Browne, School of Engineering and Science, Swinburne University of Technology
Publication Date : October 2001

Document Excerpts:

This document has been used for teaching Design for Quality course in Swinburne University of Technology, Australia. The course aim is to explore advanced quantitative methodologies for the design of mass-produced products in an environment where time-to-market and quality are critical.

Design for Quality, as it is just starting to be practised world wide, comprises a collection of analytical and experimental techniques for determining and optimizing reliabilities and making design decisions which will result in a product or process which is insensitive to tolerances and other 'noise' influences.

The methodologies are:

Probabilistic Design: How to design with parameters which have tolerances rather than just fixed values, so that the reliability of the final product coming off the end of the production line can be calculated before the design goes into production. It is covered in Chapter 2.

Robust Design: How to optimize a design to maximize its reliability, so that production may proceed using the cheapest materials and processes, without compromising the quality of the design. It is covered in Chapter 3.

Simulation: How to model the quality of thousands of units of a design coming off the production line and check that they are within specification before they are produced. How to simulate the wear processes in a machine and find out at which stage it may no longer functions satisfactorily. It is covered in Chapter 4.

This document uses Mathematica to help the formulation of problems and their solutions in Chapter 3, 4, 5. Mathematica is chosen because it has a statistics package which includes a comprehensive set of probability distributions, and it can easily do the power series expansions needed for the moment method, or the differentiations needed to approximate the variances of a function of random variables.

Arrow View/Download Probabilistic Design

ndaru
Site Admin

Joined: 09 Oct 2004
Posts: 744
View user's profileSend private message
  
   
 Reply to topic