论文部分内容阅读
工程围岩是一种高度非线性的复杂动态系统,其影响因素众多,单一的评价指标已不能准确描述围岩分类情况。目前,综合考虑多种指标评价围岩分类的方法很多,但围岩评价指标之间或多或少存在一定的相关性,其评价指标中存在一些服从非高斯分布的指标,无法满足概率神经网络(PNN)样本层中采用高斯分布作径向基函数的要求,因此,提出一种对称Alpha稳定分布(SaS)。SaS有更广泛的数学表达,其径向对称特性还可充当PNN样本层中高斯分布。在SaS的基础上,建立广州抽水蓄能电站二期工程围岩分类评价的SaS-PNN模型。预测结果表明,SaS-PNN模型具有良好的预测效果,其误判率为为4.55%。可为地下工程围岩分类评价提供一种新思路。
The engineering surrounding rock is a kind of highly nonlinear and complex dynamic system, and its influencing factors are numerous. A single evaluation index can not accurately describe the surrounding rock classification. At present, there are many ways to comprehensively evaluate the classification of surrounding rock by a variety of indexes. However, there are some correlations between the evaluation indexes of surrounding rock. The evaluation indexes have some indexes that obey the non-Gaussian distribution and can not meet the requirements of probabilistic neural network PNN) sample layer using Gaussian distribution as a radial basis function, therefore, a symmetrical Alpha stable distribution (SaS) is proposed. SaS has a broader mathematical expression and its radial symmetry can also serve as a Gaussian distribution in the PNN sample layer. On the basis of SaS, a SaS-PNN model for classification and evaluation of surrounding rock of the second phase of Guangzhou Pumped Storage Power Station was established. The prediction results show that the SaS-PNN model has a good prediction effect with a false positive rate of 4.55%. It can provide a new idea for classification and evaluation of surrounding rock in underground engineering.