سال انتشار: ۱۳۸۸

محل انتشار: سومین کنفرانس نانوساختارها

تعداد صفحات: ۵

نویسنده(ها):

Amir Rabiee Kenaree – Department of Chemical Engineering, University of Tehran, Iran
Behzad lotfi – Department of Chemistry, University of Tehran, Iran

چکیده:

Zirconia ZrO2, owing to its mechanical, electrical, thermal and optical properties and photo-chemical stability, has been the centre of attention in recent years. Due to the dependence of zirconia performance on its particle’s size, it’s important to control the size of the produced particles during the synthesis process. Since highly complex and nonlinear relationships exist between the particles’ diameters and the effective factors, a computational intelligence-based method, i.e. Radial Basis Function networks (RBF) is employed in this paper in order to predict the particles’ size in different conditions. In this process, the output variable, i.e. diameter of the nano-particles, highly depends on the concentration of the ZrOCl2.8H2O, the flow of the gas NH3 (Vs) and the level of system vacuum. Hence these three variables are chosen as the input variables. Experimental data, in two sets of training and test, have been used to train the model. The RBF network is constructed using sub-set selection technique along with the Orthogonal Least Square (OLS) learning algorithm. Statistical error criteria, including Mean Square Error (MSE), Root Mean Square of Error (RMSE), R2 are used to assess the model’s reliability. The obtained results demonstrated the capability of the RBF network in modelling, simulating and predicting the synthesis process of the c-ZrO2 nano-particles.