为了实现某型直升机可靠度准确预测,建立了基于自适应模糊神经网络的可靠度预测模型。该模型以故障间隔时间为输入,以可靠度为输出,通过模糊神经网络自适应训练获得适合直升机可靠度预测的故障间隔时间隶属函数。对模型进行检验取得了较好的仿真结果,与应用RBF神经网络进行可靠度预测相比速度更快,同时预测精度提高了98.4%。 In order to accurately predict the reliability of a helicopter, a reliability prediction model based on adaptive fuzzy neural network is established. The model takes the time between failures as input and the reliability as output. The model is adaptively trained by fuzzy neural network to obtain membership function of fault interval suitable for helicopter reliability prediction. The simulation results of the model obtained better simulation results, which are faster than the RBF neural network for reliability prediction and the prediction accuracy is improved by 98.4%.