Coal-based jet fuel is an important alternative energy source for aviation sector. To formulate the surrogate of a coal-based jet fuel, a hybrid mixing model based on explicit equations and artificial neural network (ANN) is developed to emulate fuel atomization characteristics and pollutant emissions in aero-engine combustor. Hydrogen-carbon ratio, molecular weight, and lower heating value are calculated by explicit equations, and the ANN mixing model is used to predict the density, viscosity, surface tension, and distillation curve for the surrogate mixture at various temperatures. In the ANN model, the learning task is completed through tan-sigmoidal and linear functions, and the Levenberg-Marquardt algorithm is employed for the optimization process. The resulting surrogate of the coal-based jet fuel obtained by the hybrid mixing model is composed of n-decane/ndodecane/n-tetradecane/iso-octane/methylcyclohexane (0.026/0.603/0.229/0.117/0.025 by mole). The proposed surrogate can match the physicochemical properties of the target fuel, and also shows good agreement with the target fuel in terms of atomization characteristics and CO emissions, while the NOx emissions of the surrogate is higher than those of the target fuel for most test conditions.