Tugas Akhir
Pemodelan reaktifitas doppler dari high temperature reactor menggunakan jaringan saraf tiruan
In realization of the Experimental Power Reactor which is HTGR type, it is necessary to prepare operating human resource for HTR. Kartini Reactor which is TRIGA-II type, will be part of HTR simulator by planting Doppler reactivity characteristic on it , where the simulator will be used as a training facility for licensed Experimental Power Reactor operator candidate. Doppler reactivity model based on a series of artificial neural networks (ANN) used as a process element of the HTR simulator on the Kartini Reactor, later called a hybrid simulator. Three pieces of ANN with combination of NARX and MLP architecture were trained each, before being assembled as Doppler reactivity model. The ANN model trained and validated by using PCTRAN datasets, then tested on Simulink. The result show that Doppler reactivity can be modelled by using ANN with low performance if the third input of the system is taken from ANN 3. The average MSE obtained was 0,035272.
S17-9057 | 14/TA/K/17 621.039.564.3 GAR p | Perpustakaan Poltek Nuklir | Tersedia |
Tidak tersedia versi lain