The technical analysis conducted in this study manages the modelling of diesel engine exhaust emissions using artificial neural networks. Target of this study is to comprehend the adequacy of different biodiesel fuel properties and engine working conditions on diesel engine ignition towards the development of fumes discharges. The experimental investigations have been did on a single cylinder Direct Injection (DI) combustion ignition (CI) engine utilizing mixes of biodiesel methyl esters from Cotton seed, Jatropha, Mahua, Karanja and Neem oils. The execution parameters, for example, brake power (BP), brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), exhaust gas temperature (EGT) were measured alongside managed and unregulated fumes outflows of CO, HC, NOx and smoke density. An Artificial neural network (ANN) was made in perspective of the available exploratory data. Multi layer recognition neural framework was used for nonlinear mapping amidst data and yield parameters of ANN. Biodiesel blend rate, calorific quality, thickness, Cetane number of each biodiesel blend and working burden were used as inputs to set up the neural framework. The fumes gas outflows - CO, HC, NOx and smoke thickness are foreseen for the new fuel and its mixes. Particular start limits and a couple of standards were used to get ready and endorse the institutionalized data plan and a satisfactory rate goof was expert Levenberg-Marquardt outline enhancement calculation. The results exhibited that readiness through back engendering was adequately satisfactory in suspecting the motor surges. It was found that R (Regression Coefficient) qualities were 0.95, 0.99, 0.99 and 0.98 for CO, HC, NOx and Smoke thickness outflows, individually. Along these lines, the made model can be used as a definite instrument for evaluating the emanations of biodiesels and their blends under changing working conditions.