基于黄金割理论的改进BP神经网络的高速公路安全投入评价研究

为解决高速公路安全投入构成复杂、评价定量难，以及施工企业普遍存在安全投入力度不足、不合理的问题，需要对高速公路安全投入的合理性进行评估。通过分析高速公路安全投入的构成特点，采用安全成本分解结构法（SCBS）从人员、技术设施和文明管理角度构建高速公路安全投入评价指标体系；设计基于黄金分割理论的改进BP神经网络的高速公路安全投入评价模型。以遂德高速公路安全投入评价为案例，评价结果表明该项目的安全投入比较合理。本文评价结论与该工程实际情况基本一致，验证了本文构建的高速公路安全投入评价指标体系科学合理，提出的改进BP神经网络计算方法速度和精确度较高，为高速公路安全投入合理性评价提供参考和借鉴思路。

Study on highway safety investment evaluation based on improved BP neural network based on golden section theory

In order to solve the complicated composition of expressway safety investment, difficult evaluation and quantification, and the common problems of construction enterprises, such as insufficient safety investment strength and unreasonable weight ratio.It is necessary to evaluate the rationality of expressway safety investment.By analyzing the characteristics of expressway safety investment, the evaluation index system of expressway safety investment is constructed from the perspectives of personnel, technical facilities and civilized management by using security cost breakdown structure(SCBS).And an evaluation model of expressway safety investment based on improved BP neural network based on golden section theory is designed.Taking Suide expressway safety investment evaluation as an example,the results show that the safety investment of the project is reasonable. The evaluation conclusion is basically consistent with the actual situation of the project, which verifies that the evaluation index system of expressway safety investment constructed in this paper is scientific and reasonable,the improved BP neural network calculation method has higher speed and accuracy.which provides reference and reference ideas for the rationality evaluation of expressway safety investment.