Surface roughness prediction of end milling process based on IPSO-LSSVM

Abstract

Surface roughness is a significant index in evaluating workpiece quality. So research about predicting surface roughness precisely prior to machining is necessary in order to save cost and attain high productivity levels. In this paper, a method called improved particle swarm optimization-least square support vector machine (IPSO-LSSVM) is proposed to predict the surface roughness of end milling. Firstly, an improved particle swarm optimization(IPSO) algorithm is used to optimize the parameters of LSSVM method which have significant influence on the accuracy of LSSVM model. Secondly, a surface roughness prediction model is established through LSSVM method with the optimized parameters. Then prediction accuracy of the established model can be attained through test data. Finally, the prediction accuracy of IPSO-LSSVM method is compared with the accuracy of other methods, and the results show that IPSO-LSSVM method is competent in fields of surface roughness prediction.

Keywords

• end milling
• surface roughness
• prediction
• improved particle swarm optimization(ipso)
• least square support square vector machine(lssvm)