基于粗糙集和支持向量机的深基坑工程风险评价

摘要

In this paper,we use the theory and methods of Rough Sets and Support Vector Regression machine to set up a risk assessment model based on rough sets and Support Vector Regression machine（SVR）.According to the risk evaluation example of the deep foundation pit,the reduction index is trained in the SVR.We use Gaussian RBF kernel function to construct an evaluation model.Then,we compare the evaluation result before reduction index with the evaluation results based on rough sets,and compare the evaluation result after reduction index with the evaluation results based on neural network in the model which we created in this paper.In the end,we get a conclusion that the efficiency of the evaluation method based on the support vector regression machine is better than the method of neural network.%利用粗糙集和支持向量回归机的理论和方法,建立基于粗糙集和支持向量回归机相结合的风险评价模型。结合深基坑工程风险评价的实例,将约简后的评价指标输入到支持向量回归机中训练,构建评价模型,而在模型的构建中采用了高斯径向基核函数。最后将指标约简前后所得的评价结果分别与基于粗糙集和神经网络的评价所得的结果进行比较,采用粗糙集和支持向量回归机的评价法具有更好的评价效率。