摘要 : In order to improve the curing quality of thermosetting prepreg, reduce the unevenness of the temperature field and the curing degree field during curing process, and improve the curing efficiency, a multi-objective optimization m... 展开 In order to improve the curing quality of thermosetting prepreg, reduce the unevenness of the temperature field and the curing degree field during curing process, and improve the curing efficiency, a multi-objective optimization method is used to optimize the cure cycle. In this paper, the coupling of heat conduction, cure kinetics, are used to analyze curing process of thermosetting prepreg. Firstly, a quarter finite element analysis model of the 4-layer unidirectional laminate is established in ABAQUS, the change of the cure cycle is considered, the temperature field and the curing degree field are analyzed. After comparison, the results of numerical simulation are basically consistent with the data in the reference paper. Secondly, surrogate model was established by Genetic Algorithm- Back Propagation (GA-BP) neural network, and the target value is predicted accurately under the given process parameters. The GA-BP surrogate model is used as the fitness function, and the Non-dominated sorting genetic algorithm II (NSGA-II) algorithm is used to select the maximum value of temperature overshoot and the curing time as the objectives to perform multi-objective optimization of process parameters. Finally, the research results show that the optimization method can reduce the maximum value of temperature overshoot, improve the uniformity of curing, and reduce curing time. The optimization strategy of "finite element numerical simulation-GA-BP neural network-NSGA-II optimization algorithm" is proposed, which has positive significance for the optimization of composites molding process. 收起