UNI-MB - logo
UMNIK - logo
 
(UM)
  • Prediction of steel machinability by genetic programming
    Brezočnik, Miran ; Kovačič, Miha, 1974- ; Pšeničnik, Matej
    The paper describes an intelligent system whhich was designed to predict steel machinability. It was made by means of genetic programming and the data on the batches of steel already made. The thus ... obtained mathematical model gives only 4 wrong predictions out of 146 experimental values. The model was tested on a test data set. The machinability of the complete batches (27 experimental values) was successfully predicted. The limitation of the proposed concept is the size of test data (N=27), which means a longer testing period. The 146 bateches used for modelling were collected from February 2004 to April 2005. With the proposed approach we can establish efficient planning and optimizing of production, reduce the costs of researches and the handling of changes, and finally, increase the satisfaction of customers by shorter delivery times. The presented approach is innovative and offers a high level of prediction precision. It is applicable in engineering practice.
    Source: Journal of achievements in materials and manufacturing engineering : JAMME. - ISSN 1734-8412 (Vol. 16, iss. 1/2, May-Jun. 2006, str. 107-113)
    Type of material - article, component part
    Publish date - 2006
    Language - english
    COBISS.SI-ID - 10526742

Shelf entry