Naslov
Prediction of tool wear after machining

Autori
Shahini, Filip ; Grgurić, Nikola

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the International Scientific Student Conference RI-STEM-2021 / Lorencin, Ivan ; Baressi Šegota, Sandi - Rijeka, 2021, 81-89

ISBN
978-953-8246-22-7

Skup
International Student Scientific Conference (Ri-STEM 2021)

Mjesto i datum
Rijeka, Hrvatska, 10.06.2021. - 11.06.2021

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Genetic Programming ; Regression ; Tool wear prediction

Sažetak
The purpose of this article is to regress and predict tool wear from predefined datasets. Given datasets consists of six different parameters: experiment number, depth of cut (DOC, mm), cutting speed (CS, m/min), feed rate (FR, mm/rev), tool wear (μm), surface roughness (μm). The goal is to merge measurements, eliminate unnecessary elements from datasets, use genetic programming method to regress the value of the "tool wear" column, predict values based on the given dataset and to evaluate regress quality using R2 and RMSE metrics. Many of the python libraries were used to successfully solve a given problems. To point out, pandas for processing given datasets, gplearn for genetic programming algorithm, sklearn for metrics evaluation and split test and train samples and matplotlib for graphical representation of the result. In addition, comparison to Random Forest Regressor algorithm and Decision Tree Regressor algorithm was made in order to confirm the results. The conclusion of this research is that the two given datasets were not compatible for merging into one measurement so the regression and tool wear prediction were not usable.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Temeljne tehničke znanosti, Interdisciplinarne tehničke znanosti

POVEZANOST RADA