Inteligentni nadzor obrabe rezalnega orodja s spremljanjem toka na glavnem vretenu

magistrsko delo

Niko Turšič (Author), Riko Šafarič (Mentor), Simon Klančnik (Mentor)

Abstract

V magistrski nalogi je predstavljen sistem za nadzor rezalnega orodja, ki temelji na sledenju toka na glavnem vretenu z uporabo umetne nevronske mreže. Glavni namen aplikacije je razširiti vpogled, ki ga ima operater v stanje orodja med delovanjem stružnice CNC. Program za analizo lahko nemoteno deluje paralelno na procesnem računalniku in prejema podatke preko podatkovnega omrežja s krmilnika stružnice. V delu so predstavljeni proces zajemanja podatkov za učno bazo umetne nevronske mreže tipa Long-Short Term Memory, arhitektura in učenje nevronske mreže, ki je uporabljena v tej aplikaciji, ter validacija naučenega modela z umetno inteligenco na novih podatkih. Prav tako sta predstavljena tudi izdelava in delovanje programa, ki se lahko izvaja na procesnem računalniku za potrebe pomožne diagnostike orodja.

Keywords

tok na glavnem vretenu;nevronska mreža;nadzor obrabe orodja;LSTM;umetna inteligenca;magistrske naloge;

Data

Language:	Slovenian
Year of publishing:	2023
Typology:	2.09 - Master's Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[N. Turšič]
UDC:	621.941.025:004.8(043.2)
COBISS:	172751619
Views:	101
Downloads:	16
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Intelligent tool condition monitoring system utilizing spindle current measurements
Secondary abstract:	The thesis presents a tool condition monitoring system, based on the tracking of current flowing through the main spindle with an artificial neural network. The main purpose of this application is to provide additional insight to the operator regarding tool wear during the operation of the CNC lathe. The TCM program can run independently on a process computer next to the lathe, receiving data via the ethernet network from the machines PLC. We present the measurement processes with which we have obtained the training data for the Long-Short Term Memory neural network, the design and training of said network and the validation of the artificial intelligence model on a new dataset. Along with the trained model we also provide a prototype software designed to use said model for the purposes of assistive tool condition monitoring.
Secondary keywords:	main spindle current;neural network;tool condition monitoring;long-short term memory;artificial intelligence;
Type (COBISS):	Master's thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Mehatronika
Pages:	1 spletni vir (1 datoteka PDF (VIII, 61 f.))
ID:	19803827