home » groups » IPROMS 2006 » IPROMS 2006 » Decision Support

Prediction of workpiece surface roughness using soft computing

A study is presented to model surface roughness in end milling using adaptive neuro-fuzzy inference system (ANFIS).The machining parameters, namely, the spindle speed, feed rate and depth of cut have been used as inputs to model the workpiece surface roughness. The parameters of membership functions (MFs) have been tuned using the training data in ANFIS maximizing the modeling accuracy. The trained ANFIS are tested using the set of validation data. The effects of different machining parameters and number of MFs on the prediction accuracy have been studied. The procedure is illustrated using the experimental data of end-milling 6061 aluminum alloy. Results are compared with artificial neural network (ANN) and previously published results. The results show the effectiveness of the proposed approach in modeling the surface roughness.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

NOTE: TO READ THIS AND OTHER IPROMS 2006 PAPERS, PLEASE REGISTER FOR THE CONFERENCE.

REGISTRATION IS FREE.

CLICK here TO REGISTER.

‹ Statistical approach to numerical databases:clustering using normalised Minkowski metricsOptimization of Assembly Lines with Transportation Delay Using IPA ›
a pdf file
Submitted by ashraf_afify on Thu, 29/06/2006 - 10:27am.
login or register to post comments |
Varying number of input features
Submitted by charles on Mon, 03/07/2006 - 12:42pm.

Hi,

An interesting application of prediction with ANFIS. Would you be able to give a bit more detail regarding varying the number of input features. Particularly the reson behind this and the effect with individual features.

Thanks,
Charles.

login or register to post comments |
Varying number of input features
Submitted by samantab on Tue, 11/07/2006 - 7:50pm.

Hi Charles,

Thanks for your comments and the question. The number of input features was varied to see their relative effects on the prediction accuracy of surface roughness. In the present case, all 3 inputs with 3 MFs gave the best results. The first 2 inputs (speed and feed) together also gave satisfactory results. In the present study, there were only 3 inputs - so the selection of most pertinet features can be done easily. In cases where there are several inputs (more than 5 or so), the selection of suitable inputs may be necessary for computational speed and modeling accuracy.

Thanking you once again for your interest and sorry for the dealy in my reply due to other preoccupations.

Best regards,

samanta

login or register to post comments |
Feature selection
Submitted by charles on Wed, 12/07/2006 - 9:16am.

Hi Samanta,

You are quite right, feature selection is the important and difficult task in this kind of applications.

Thanks,
Charles.

login or register to post comments |

Comment viewing options

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

User login

Captcha Image: you will need to recognize the text in it.
Please type in the letters/numbers that are shown in the image above.