A study was conducted for fuel characterization of a stable nano-oxide added water emulsified biodiesel blend (NWEB) prepared with optimal emulsifying parameters. NWEB prepared with Hydrophilic-lipophilic balance (HLB) value 5.0 indicated the least separation from emulsion layer as compared to other HLB values 4.3 and 6.0. Fourier transformed infrared spectroscopy was conducted in Mid-IR range to ensure the quality of biodiesel and emulsified fuel. The stability data of NWEB prepared with varying the emulsifying parameters were used in radial basis function neural network (RBFNN) for simulation of actual and predicted separation from emulsion. The well trained RBFNN model was coupled with particle swarm optimization (PSO) technique to find the optimal emulsifying parameters to produce a stable NWEB. The predicted optimal emulsifying parameters were found as 69.7 ppm nano-oxide concentration, 10% water, 1% surfactant and 2500 rpm of stirrer for a predicted separation of 0.90% after 30 h of emulsification against actual separation of 0.84%. The fuel properties of NWEB and base fuels were determined following the standard procedures. These properties were found within the limits of American, European and Indian Standards for diesel, biodiesel and biodiesel blends and thus indicating the suitability of these fuels in diesel engine. (C) 2019 Elsevier Ltd. All rights reserved.