【摘要】 应用人工神经网络对舰炮发生故障的发射炮弹数进行了预测。借助前馈神经网络对非线性函数的逼近能力 ,对发生故障的发射炮弹数进行了单步预测 ,并与使用灰色数列预测的结果进行了对比 ,验证了人工神经网络进行可靠性预测的可行性更多还原

【Abstract】 In this paper, ANN was used to predict the failure number of naval ordnance. With the nonlinear approach ability of Back Propagation Neural Network, single_step prediction of the failure number of naval ordnance was made. Comparison with the result of grey serial forecast was made, and the feasibility was proved that artifical neural network can be used for reliability prediction.更多还原