Fuzzy Logic Modelling of a Powder Coating System

The successful application of powder coating technology depends on resolving practical issues such as reducing overspray and maximising transfer efficiency. Maximising transfer efficiency requires optimising application parameters that influence powder deposition efficiency. This paper presents the design and development of a predictive fuzzy-logic model for powder coating applications. To improve the performance of the fuzzy-logic model, the model was design based on optimised conditions as determined by statistical analysis. The fuzzy-logic based powder coating system has four application variables as inputs and one output, and gives excellent performance under the conditions investigated. The fuzzy-logic model can easily be implemented in the factory and has the ability significantly to reduce waste while maximising first-pass transfer efficiency.


gbecci's picture
Submitted by gbecci on Tue, 05/07/2005 - 12:39pm.

1) What are the criteria for selecting triangular Membership Function and Mamdani Type Model?

 

2) What hardware platform was used?


Pham's picture
Submitted by Pham on Sun, 10/07/2005 - 6:31pm.

Thank you for your interest in our paper.

We used triangular MFs because of their simplicity.  We prefer the Mamdani model for the readability of the rules.

The modelling work was performed on an ordinary Pentium PC.

 


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Who's online

There are currently 0 users and 119 guests online.