高速公路路堑高边坡工程施工安全越来越受到重视,与之相关的施工安全风险评估制度被建立起来。现行的指南对评估工作做出了详细的规范与引导,但是通过实践发现被推荐使用的指标体系法存在两点不确定性,导致评估工作进度缓慢、可信度不高。该文提出了使用智能算法——神经网络法代替指标体系法,并以云南某高速公路为例,用神经网络工具箱中神经网络拟合(Neural Net Fitting)程序包进行评估,该法操作简便,计算速度快,结果与指标体系法比较更为严谨,加快了评估工作的进度。 Construction safety of highway slope cutting slope is paid more and more attention, and related construction safety risk assessment system has been established. The current guidelines provide detailed norms and guidance for the assessment work. However, it is found that there are two uncertainties in the recommended index system through practice, which leads to the slow progress of evaluation work and the low credibility. This paper proposes to use the intelligent algorithm - neural network instead of the index system method. Taking an expressway in Yunnan Province as an example, this paper evaluates the Neural Net Fitting package in the neural network toolbox, which is easy to operate , The calculation speed is fast, the result is more rigorous compared with the index system method, and the progress of the assessment work is accelerated.