BP神经网络的身管寿命预测方法

Study on Prediction Method of Barrel Life based on BP Neural Network

作为影响火炮寿命的炮膛烧蚀和磨损,其作用过程是一个相当复杂的综合过程。到目前为止,还不能建立起一个较好的数学模型来计算火炮内膛烧蚀磨损量。针对这种情况,在实测数据的基础上,提出了采用BP(Back Propagation)神经网络的方法,利用其优越的非线性逼近能力和泛化能力来计算炮膛烧蚀磨损量,并根据最大烧蚀磨损量进行身管寿命预测。

Bore erosion and wear can affect the life of gun and their action process is a complicated integrated process. Until now, a good mathematic model for computation of gun bore wear figure is not established. Based on the data measured in the practice, the paper proposes a method of calculating bore erosion and wear figures using its nonlinear approximation and generalization abilities based on BP （Back propagation） neural network. And prediction of barrel life is carried out based on maximum erosion and wear figures.