Title :
Neuro-fuzzy state variables estimators of a two-mass drive system
Author :
Kaminski, Marcin ; Szabat, Krzysztof
Author_Institution :
Fac. of Electr. Eng., Wroclaw Univ. of Technol., Wroclaw, Poland
Abstract :
In this paper neuro-fuzzy state variable estimators of drive system with an elastic coupling are presented. The described case is related to calculating the load speed and shaft torque based on easy to measure signals. During the design process, fuzzy C-means clustering was applied in order to optimize the structure of tested estimators. The obtained results show very high precision of estimation of both state variables for a wide range of reference speed and load torque changes. Also, tests for the changing mechanical time constant are presented. Also in this case the estimation is accurate. Simulations were verified in experiment prepared using dSPACE1104.
Keywords :
drives; fuzzy set theory; mechanical engineering computing; pattern clustering; signal processing; state estimation; dSPACE1104; elastic coupling; fuzzy C-means clustering; load speed; mechanical time constant; neuro-fuzzy state variables estimators; shaft torque; two-mass drive system; Estimation; Fuzzy systems; Load modeling; Mathematical model; Shafts; Torque; Transient analysis; load speed; neuro-fuzzy models; shaft torque; state variables estimation; two-mass system;
Conference_Titel :
Industrial Technology (ICIT), 2015 IEEE International Conference on
DOI :
10.1109/ICIT.2015.7125346
