摘要
施工振动风险评估是以大数据为基础的,风险评估的结果受很多因素的影响。为了从诸多影响因素中提取关键影响因素并准确评估施工振动风险。提出了一种LLE非线性特征提取与ANFIS评估相结合的风险评估方法。该方法首先利用LLE流形方法提取众多施工振动风险评估样本影响因素组成的高维数据向量的非线性成分,然后将该非线性成分作为ANFIS评估方法的输入对施工振动风险样本进行评估分类。经实例验证该方法可以有效提高风险评估的准确性,降低错分率。将其应用到实际工程中能为施工管理者提供有效决策信息,提高决策的可信度。
The risk assessment of construction vibration is based on big data,and the result of risk assessment is affected by many factors.In order to extract key influencing factors from many influencing factors and accurately assess the risk of construction vibration.A risk assessment method combining LLE nonlinear feature extraction with ANFIS evaluation is proposed.First the LLE manifold method is used to extract the nonlinear components of the high dimensional data vectors composed of the factors affecting the vibration risk assessment samples.Then the nonlinear component is used as the input of the ANFIS evaluation method to classify the construction vibration risk samples.It is shown that through the example that the accuracy of risk assessment can be effectively improved and the misclassification rate can also be reduced.
引文
[1]黄娟,余俊,雷明锋,等.成槽冲击荷载对邻近文物建筑的振动影响分析[J].铁道科学与工程学报,2018,15(5):1164-1170.
[2]王鑫,韩煊,周宏磊.中心城区地铁盾构施工引起的环境振动响应研究[J].土木工程学报,2015,48(S2):309-314.
[3]张涛,吕淑然,杨凯.强夯振动对周边环境的影响研究[J].中国安全科学学报,2014,24(11):127-132.
[4]刘东海,高雷.基于碾振性态的土石坝料压实质量监测指标分析与改进[J].水力发电学报,2018,37(4):111-128.
[5]Hao H,Ang T C,Shen J.Building vibration to traffic-induced ground motion[J].Building and Enviroment,2001,36(3):321-336.
[6]Ebadi M M,Doudak G,Smith I.Vibration responses of glulam beam-and-deck floors[J].Engineering Structures,2018,156:235-242.
[7]王鑫,韩煊.盾构施工振动振源的影响因素研究[J].地震工程学报,2014,36(3):592-598.
[8]王祥秋,周岳峰,周治国.爆破冲击荷载下隧道振动特性与安全性评价研究[J].中国安全科学学报,2010,20(11):134-138.
[9]赵博,赵亚琼.边坡稳定性能的判定方法研究及其工程应用[J].水利与建筑工程学报,2017,15(1):25-29.
[10]王衍汇,倪万魁,石博溢,等.延安新区黄土高填方边坡稳定性分析[J].水利与建筑工程学报,2014,12(5):52-56.
[11]Ngamkhanong C,Kaewunruen S.The effect of ground borne vibrations from high speed train on overhead line equipment(OHLE)structure considering soil-structure interaction[J].Science of The Total Environment,2018,627:934-941.
[12]丁少梅,赵忠虎,王宁宁,等.岩石强度理论的分类评述[J].水利与建筑工程学报,2017,15(1):95-102.
[13]Hettiarachchi R,Peters J F.Multi-manifold LLE learning in pattern recognition[J].Pattern Recognition,2015,48(9):2947-2960.
[14]Bülent Haznedar,Adem Kalinli.Training ANFIS structure using simulated annealing algorithm for dynamic systems identification[J].Neurocomputing,2018,302:66-74.
[15]Yaseen Z M,Ebtehaj I,Bonakdari H,et al.Novel approach for streamflow forecasting using a hybrid ANFIS-FFA model[J].Journal of Hydrology,2017,554:263-276.