Abstract: In order to solve the problem of lag in rejection caused by the lack of real-time quality feedback of the die-casting system, a casting quality prediction method based on genetic algorithm back propagation (GABP) neural network was proposed, taking the die-casting system of the outer water jacket of the motor as the research object. By analyzing the mold temperature control mechanism, the input temperature parameter of mold-temperature controlling device was used to indirectly characterize the mold temperature change trend, and the temperature data was fitted with the curve combined with the third-order Fourier series to characterize the dynamic change characteristics of the temperature parameters of the die-casting cycle. Based on the third-order Fourier series temperature sample data, the GABP neural network model was used for training, the quality prediction model of die castings was constructed, and the weights and thresholds of the backpropagation neural network were optimized by genetic algorithm, and the casting quality prediction results were embedded into the digital twin model of the die-casting system, so as to realize the real-time monitoring and feedback of the quality of die castings.