摘要
民机交易价格高低直接影响着交易双方的利益,交易价格估算方法已经成为确定价格的支持工具,越来越受到重视。当前,对民机交易价格估算模型的研究并不多,而且一般局限于对新机价格或者目录价格估算。鉴于此,综合考虑影响交易价格参数,如飞机几何特征、性能特征、机龄等作为输入参数,构建了一种基于广义回归神经网络的民机交易价格估算模型。并使用行业交易样本数据对模型进行了训练,寻找到最优参数后,用其进行重构模型及模型估算精度验证。实例证明,模型具有良好泛化能力和估算精度高等特点,可以用于民机交易价格估算与预测。
There is now a prior focus on the estimation model of civil airplanes transaction price,which is an important component in decision support system aiding to determine the actual price of negotiations. Because the transaction price will directly affects the interests of both parties. At present, there are not many investigations on civil airplanes transaction price estimation model, and generally confined to the estimation for new price or list price. In view of this, in this paper, a civil airplanes transaction price estimation model based on general regression neural network is established considering the influences as inputs, such as airplane geometry, performance characteristics and age. The model is trained withsample trade data, then model is reconstructed with the found optimal factor, and finally the ability and accuracy of estimation model is verified. Examples show that the model has good generalization ability and high estimation accuracy, and it can be used for civil aircraft price estimation and prediction.
出处
《航空计算技术》
2017年第6期75-79,共5页
Aeronautical Computing Technique
基金
航空科学基金项目资助(2016ZA53002)
江苏省高校优势学科建设工程项目资助
