片岩隧道变形规律分析与预测方法研究

为准确掌握片岩隧道变形规律,基于隧道变形监测结果,利用核极限学习机构建隧道变形初步预测模型,通过遗传算法和蚁群算法进行优化处理,以保证模型参数的最优性,采用混沌理论对预测误差进行修正处理,利用M-K分析判断隧道变形趋势,并将趋势判断结果与预测结果对比.结果表明:通过递进优化处理,能逐步提高预测精度,且预测结果的相对误差...

Study on deformation law analysis and prediction method of schist tunnel

In order to accurately grasp the deformation law of schist tunnel,based on the results of tunnel deformation monitoring,the kernel limit learning machine was used to build a preliminary prediction model of tunnel deformation,and the optimization processing of genetic algorithm and ant colony algorithm were used to ensure the optimality of parameters in the model.The chaos theory was used to correct the prediction error,then the M-K analysis was used to judge the trend of tunnel deformation,and the trend judgment results were compared with the prediction results,so as to realize the comprehensive study of tunnel deformation law.The results showed that the prediction accuracy could be improved step by step through the progressive optimization processing,and the average relative error of the prediction results was between 1.77%~1.83%,which were all less than 2%,so the prediction accuracy was higher.At the same time,by judging the deformation trend,it was found that although the tunnel deformation had an increasing trend,the trend continued to weaken with time.Therefore,the tunnel deformation presented an increasing trend with small rate and tended to be stable generally.