为提高燃气轮机研制过程中的风险管理能力，针对果蝇算法（Fruit Fly Optimization Algorithm，FOA）及BP神经网络的缺陷，构建了自适应果蝇算法（Adaptive Fruit Fly Optimization Algorithm，AFOA），提出基于自适应果蝇算法优化BP神经网络的风险预测模型，利用自适应果蝇算法优化BP神经网络的阈值和权值。挖掘燃气轮机研制风险因素及风险事件之间的关系，并根据风险因素的权重预测风险事件的权重。利用燃气轮机研制风险的相关历史数据进行验证，表明该模型具有较高的预测精度和应用价值。

In order to improve the ability of risk management in the process of gas turbine development,aiming at the inherent defects of fruit fly optimization algorithm (FOA) and BP neural network,the adaptive fruit fly optimization algorithm (AFOA) is built,and the risk prediction model based on AFOA optimized BP neural network is proposed,by use of the parameters of the thresholds and weights of BP neural network optimized by AFOA.The prediction model can explore the relationship between risk factors and risk events in gas turbine development,and predict the weights of risk events according to the weights of risk factors.The model is verified by the relevant historical data about gas turbine development risk.The result shows that the model has higher prediction accuracy and application value.

C935

国家科技重大专项（2017-Ⅰ-0011-0012）；国家自然科学基金委项目（51705436）；四川省科技计划项目（2021JDRC0174）；船舶与海洋工程动力系统国家工程实验室-海洋工程燃气轮机实验室资助项目