Deformation Sign Monitoring Method Of Concrete Gravity Dam Based On The Excavation Of The Multivariate Space-time Information Mining

The existing concrete dams of China rank first in the world both in terms of quantity and scale,and those long-term health has an critical impact on the safety of the entire hydraulic projects.Due to the concrete dam is affected by the external environmental load and the internal material performance evolution during the operation period,the performance of dam structure will be degraded to varying degrees.Establishing safety monitoring models based on the dam prototype monitoring data,mastering and interpreting the operational performance of the dam structure and ensuring the dam safety are important scientific issues that should be concerned.Therefore,takes the gravity dam as the object in the article,and the research on the operation monitoring and deformation sign control method of concrete gravity dam is carried out by the theoretical analysis,numerical simulation and prototype observation.The main contents are as follows:(1)Considering that there are inevitably loss and anomaly in the prototype monitoring data,in order to further enhance the ability of the monitoring data interpreting the operational performance of the dam,combined with kernel independent principal component analysis and extreme learning machine,a missing data interpolation model for multi-source spatiotemporal information is constructed.In addition,the singularity diagnosis method of monitoring data is proposed by using singular spectrum analysis and 3? criterion.(2)With the increase of the operating years,the dam materials inevitably have different degrees of aging and performance degradation.The dam structural material parameters have changed to some extent compared with the design period.Considering that the monitoring data has a certain error in the actual engineering field,the structural analysis results based on the deterministic structural parameters and the physical model are inevitably distorted.Therefore,the interval inversion model of the material parameters of the concrete dam is established based on the rough set theory and the neural network principle.(3)Because the function between deformation and environmental factors in the traditional monitoring model is not fully understood,the function between deformation and sedimentation in the reservoir area is derived,and a correlation vector machine based on flock algorithm optimization is proposed to construct a statistical model of deformation monitoring.In addition,based on the material parameters inversion,a corresponding deformation monitoring hybrid model is constructed to better relate the structural evolution characteristics of the dam system.(4)The early warning indicator of deformation is an important indicator,which can feedback and control the operational behavior of the dam.On the basis of introducing the typical small probability method and the mixed model method to formulate the deformation monitoring index method,the monitoring index of the horizontal displacement of the dam crest is drafted by the aforementioned deformation monitoring positive and negative analysis results.