With the importance of environmental issues and the increasing reduction of fossil fuel resources, optimizing fuel consumption and reducing the concentration of combustion pollutants has become one of the most important concerns of researchers in this field. Therefore, prediction of temperature and concentration of combustion products has been the main goal of many researchers in recent years. Due to the importance of the bluff-body flame and its application in industry, this type of flame was investigated in the current work. In this study, the chemical kinetics mechanism of GRI-Mech3.0 was used and after generation of flamelet library using Fluent, Cantera, and Chemkin, it was observed that the answers obtained from Cantera and Chemkin are very similar to each other, while the results of Fluent flamelet are different from other softwares. After generating flamelet library, by applying two probability density functions of beta and clipped Gaussian and generating a look-up table, the performance of these two functions for the adiabatic bluff-body flame was investigated using the flamelet model and it was concluded that the beta function predicts the temperature distribution and mass fraction of species more accurately than the clipped Gaussian function. After selecting the beta function as an appropriate probability density function for the mixture fraction variable, the adiabatic bluff-body flame was simulated using the FPV combustion model. For this simulation, first, a set of data with three inputs of mixture fraction, mixture fraction variance, and reaction progress variable and output of scalar dissipation rate were extracted using a preprocessing program and the Cantera flamelet, and then an artificial neural network was built on this data and after solving the neural network, the weight and bias coefficients of this network were given to the main code and the flame was simulated using the FPV model. In this simulation, to solve the temperature field, two methods of solving the energy equation and interpolation of the look-up table were used, but no significant difference was observed in the results of these two methods. By comparing the results with experimental data, it was observed that in both methods of temperature estimation, the results of the FPV model are more accurate than the results of the flamelet model. Keywords : Flamelet model, FPV model, Non-Premixed flame, Bluff-Body flame, Probability Density Function, Artificial Neural Network, Flamelet library