基于BP神经网络电子评标方法的研究和应用  

Evaluation method based on BP neural network in tendering

作　　者：宋早雪[1] 周丽萍[1] 杨毅[1] 韩爱红[1] 

机构地区：[1]西北工业大学力学与土木建筑学院,陕西西安710072

出　　处：《南京工业大学学报（自然科学版）》2007年第4期85-89,共5页Journal of Nanjing Tech University(Natural Science Edition)

基　　金：河南省科技攻关计划基金资助项目(496450009);西北工业大学研究生创业种子基金资助项目(Z200624)

摘　　要：为了解决招投标中评标环节专家评审法随意性大的问题,针对建设工程的不同特点,对评标内容进行分类剖析.利用人工神经网络的智能识别功能,建立基于神经网络的电子评标方法.结合人工神经网络的自适应功能、良好的容错能力以及非线性数据处理特点,科学地评价多因素建设工程投标方案,使招标商从众多的投标商中选出最佳的中标单位.In order to solve the problem of arbitrariness when the expert evaluates the tenders, and considering differences between the construction projects, the tender evaluation contents were classified and analyzed. By intelligence distinguishing function of manpower neural network, the neural network-based electronic tender evaluation method was established. With manpower neural network self-adaptive functiions, fault-tolerant ability and nonlinear data handling characteristics, the possibility to scientifically evaluate the multi-factor construction project tender schemes was obtained. The method assures to select out the optimal tendering company from all tenderings.

关 键 词：招投标 评标方法 人工神经网络 BP算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]