Filtration of Estimated Flank Wear Widths Using Recurrent Neural Network in the Tool Wear Regulation Process (CROSBI ID 569560)
Prilog sa skupa u zborniku | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Brezak, Danko ; Majetić, Dubravko ; Udiljak, Toma ; Staroveški, Tomislav
engleski
Filtration of Estimated Flank Wear Widths Using Recurrent Neural Network in the Tool Wear Regulation Process
In order to accomplish robust and reliable tool wear regulation it is necessary that the filter of estimated tool wear parameters is implemented in the control loop. In this research, the proposed Modified Dynamic Neural Network filter with hidden layer neurons built in the form of Dynamic Elementary Processor (DEP) is analysed. Different configurations of network structure with DEP units structured as an IIR and FIR filters are tested separately and in the control loop using two intervals of estimation errors of the flank wear parameter and ‘impulse’ process disturbances.
Control; Tool Wear; Neural Network
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
104-107.
2010.
objavljeno
Podaci o matičnoj publikaciji
Proceedings of the CIRP ICME '10 International Conference
Teti, Roberto
Napulj:
978-88-95028-65-1
Podaci o skupu
CIRP ICME ’10 - 7th CIRP International Conference on Intelligent Computation in Manufacturing Engineering
predavanje
23.06.2010-25.06.2010
Capri, Italija