AHP方法和BP神经网络在装备保障性评估中的应用  

AHP and BP Neural Network's Application in Equipment Indemnificatory Valuation

作　　者：李兵[1] 郭煌[1] 郭建胜[1] 惠晓滨[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《西安航空技术高等专科学校学报》2007年第5期23-26,31,共5页Journal of Xi'an Aerotechnical College

摘　　要：目的介绍了BP神经网络的基本原理,建立装备保障性评估指标体系。方法在系统分析装备保障性的各种影响因素的基础上,设计用于评估装备保障性的神经网络模型,利用AHP方法确定指标的初始权重,并调用MATLAB神经网络工具箱中的GUI对结果进行计算和分析。结果研究表明神经网络具有很强的解决复杂非线性关系问题的特点,适用于对装备保障性的分析和评估。结论BP神经网络评估方法是一种基于非线性系统的预测方法,对装备保障性进行综合评估能够较为准确地提供决策依据。Object： The essay introduced basic principles of BP neural network and established valuation system on equipment indemnificatory. Approach： Designing neural network model uses in equipment indemnificatory valuation on the basis of systematically analyzed various influential factors and identifying initial assessment with AHP and assigned GUI in MATLAB to compute and analyze the result. Outcome： The research suggest that neural network could work with characteristics of solving complicated non - linear to adapt to analysis and assessment on equipment indemnificatory. Conclusion： BP neural network assessment approach proves a predicting approach based on non- liner system, could accurately provide decisive evidence for comprehensive assessment in equipment indemnificatory.

关 键 词：BP网络 AHP 装备保障性 评估 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]