7050 aluminium alloy, as an important lightweight high-strength structural material because of its excellent characteristics, has increasing applications in aviation, aerospace, and launch vehicle manufacturing. Fatigue failure accounts for approximately 80 % of the structural failures in the engineering field, which leads to numerous losses and is highly uncertain and sudden. Therefore, fatigue life prediction of 7050 aluminium alloy has become a prominent research field. In this study, considering the two variables of stress ratio and stress concentration coefficient, fatigue tests of 7050 aluminium alloy under five different working conditions were conducted to evaluate the effect of these factors on the durability of 7050 aluminium alloy. A novel hybrid neural network model was proposed to solve the time-consuming problem of obtaining a large amount of S-N curve data through traditional fatigue tests. The model uses relatively low cycle fatigue test data for training, realizes data derivation, accurately predicts the fatigue limit of materials, and realizes the purpose of quickly evaluating the fatigue properties of 7050 aluminium alloy materials under different working conditions.