基于BP神经网络的飞机机体研制费用估算

Development Cost Estimation of Aircraft Frame Based on BP Neural Networks

提出一种应用人工神经网络预测飞机机体研制费用的方法。该方法利用偏最小二乘法对飞机性能参数进行了主成分提取,用这些主成分作为 BP 网络的输入变量,建立了一种基于 BP 神经网络的飞机研制费用的估算模型,应用该模型对典型的飞机机体研制费用进行了预测。预测结果表明,该方法是有效的、可行的。

A forecasting method of artificial neural network is proposed to estimate the development costs of aircraft frame.The method selects primary components of the aircraft performance parameters by the partial least square re- gression,and regards them as the BP network's input variables to build a development cost forecasting model based on BP network.The model is applied in development cost forecasting of typical aircraft frames,the results forecasted show that the method is valid and feasible.