山岭公路隧道施工安全风险评估方法及其应用研究

STUDY ON SAFETY RISK ASSESSMENT METHOD OF MOUNTAIN RIDGE HIGHWAY TUNNEL CONSTRUCTION AND ITS APPLICATION

由于山岭公路隧道施工安全风险大小与各安全风险影响因子之间存在非线性映射关系,引入混合核函数和粒子群算法,建立基于粒子群算法优化混合核极限学习机的山岭公路隧道施工安全风险评估方法。该方法依据山岭公路隧道地质与施工特征,建立隐含层节点特征映射混合核函数,采用粒子群算法对训练样本进行模型训练,同时对映射函数中混合核函数参数进行最小均方差寻优,获得山岭公路隧道施工安全风险评估的混合核极限学习鳩型并进行算例验证,结果表明所建模型可靠。

As a result of the nonlinear mapping between construction safety risk of mountain ridge highway tunnel and influencing factors of safety risk,the hybrid kernel function and particle swarm optimization algorithm are applied to establish safety risk evaluation method of mountain ridge highway tunnel construction based on the hybrid kernel limit learning machine optimized by particle swarm optimization algorithm.According to the geological and construction characteristics of mountain ridge highway tunnel,this method establishes the hybrid kernel function of hidden layer node feature mapping,uses particle swarm optimization algorithm to train the training samples,and optimizes the parameters of the hybrid kernel function in the mapping function to the minimum mean square deviation,finally obtains the hybrid kernel limit learning machine model of mountain ridge highway tunnel construction safety risk assessment,and carries out the example verification.The results show that the model is reliable.