结合灰色理论与BP神经网络建立了灰色神经网络组合预测模型.该模型有效地将灰色理论弱化数据序列波动性的优点和神经网络特有的非线性适应性信息处理能力相融合.利用三种灰色模型及灰色神经网络模型对某型空舰导弹某电子设备的失效时间进行预测,研究结果表明灰色神经网络预测效果明显优于灰色模型,是一种实用的且精度较高的预测方法. Combined with gray theory and BP neural network, a combined forecasting model of gray neural network is established. This model effectively combines the advantages of gray theory weakening data sequence volatility and neural network-specific nonlinear adaptive information processing capability. By using three gray models And gray neural network model to predict the failure time of some electronic equipment of a certain type of ship-to-ship missile. The results show that the gray neural network forecasting effect is obviously superior to the gray model, which is a practical and accurate forecasting method.