基于BP神经网络及熵权灰靶法的分包商选择研究

Research on subcontractor selection based on BP neural network and entropy grey target method

在建筑工程分包专业化发展的趋势下,项目分包商的选择变得尤为重要,分包商选择不当会给施工带来更大的风险.同时分包商招标时也存在着控制价难以确定、评价方法单一等问题为此提出了基于BP神经网络和熵权灰靶法的两阶段分包商选择模型首先提出了转换值概念,找到了承包价与分包价之间的转换关系,以S建筑公司的以往工程数据为基础运用BP神经网络预测分包招标控制价然后建立综合指标评价体系,利用熵权法确定指标权重,再用灰靶决策确定最优分包商最后以S建筑公司的实际的分包过程为例,验证了该选择模型的可行性为分包商的选择提供了新思路.

With the development of subcontractor specialization,the choice of subcontractor becomes more and more important.The improper choice of subcontractor will bring more risks to construction.At the same time,subcontractors also have problems such as difficulty in determining control price and single evaluation method when bidding.Therefore,a two stage subcontractor selection model based on BP neural network and entropy weight grey target method is proposed.Firstly,the conversion value algorithm is proposed,and the conversion relationship between the contract price and the subcontract price is found.Based on the previous engineering data of S Construction Company,BP neural network is used to predict the subcontract tender control price.Then the comprehensive index evaluation system is established,the index weight is determined by entropy weight method,and the optimal subcontractor is determined by grey target decision.Finally,taking the actual subcontracting process of S Construction Company as an example,the feasibility of the selection model is verified.It provides new ideas for subcontractors’ choice.