Research On Cluster Analysis Of Dam’s Deformation Displacement Intensity Based On Gene Expression Programming Algorithm

With the development of dam deformation monitoring technology, the acquisition of damdeformation monitoring data is becoming more and more diversified. But the data analysismethod is mainly based on the forecast analysis on a single point, can’t reflect the overalldeformation law of dam. People urgently need a method that can quickly analysis the overalldam deformation law. Therefore, through cluster analysis for of dam displacement Intensity tofind out the whole deformation law of dam has important significance. For these problems indam deformation monitoring data analysis, In this paper, mainly research the dam deformationmonitoring data analysis, and applied the theory and research methods of other areas to damdeformation monitoring data analysis, the main purpose in this paper is puts forward a methodthat can quickly analyze the overall deformation law of dam from the mass of dam deformationmonitoring data. The main research work as follows:(1) Make a deep analysis to the displacement intensity; combining the characteristicsof the dam deformation monitoring data, select all the dam deformation monitoring in theaverage displacement rate of the monitoring period as the critical speed, and as a benchmarkto determine the intensity of displacement value of dam;(2) According to the mutation noise of the dam deformation monitoring data mayproduce adverse effect of clustering analysis, the wavelet threshold denoising algorithm isadopted to the original dam deformation monitoring data for noise removal;(3) In order to eliminate the blind spot in dam overall deformation law analysis,resultfrom uneven distribution of dam monitoring points and larger space interval between dammonitoring points, generate1m*1m uniform space grid, using radial basis interpolationalgorithm to interpolate the grid displacement intensity.(4) According to the shortcoming existing gene expression automatic clusteringalgorithm in high dimension space data, improved the existing algorithm. Proposed geneexpression automatic clustering algorithm based on principal component, and map the damdeformation monitoring data of high dimension to low dimension space, through dimensionreduction to realize the dam deformation monitoring data clustering operation;(5) Using.NET mixed with the Matlab programming technology, realize the improved gene expression programming of dam displacement intensity clustering analysismodel;(6) Applied gene expression programming of dam displacement intensity clusteringanalysis model to Shangyou reservoir dam deformation monitoring projects that located inganzhou city in Jiangxi province, and analysis the whole deformation law of this dam.Through dam displacement intensity clustering analysis can clearly see that the locationof river downstream of the reservoir dam deformation on activity is stable, the location of theupstream deformation activity more acute, middle ranks between deformation activities; Theleft bank of the dam deformation is stable, the right bank deformation activities are morefrequently; And in the direction of flow deformation activity superior to the dam axis and thedirection of vertical displacement. Compare the cluster visualization rendering anddisplacement intensity in all directions,We can found that the results of clustering and thedeformation of the dam is basically consistent, which proves that the dam displacementintensity clustering analysis model based on gene expression programming and principalcomponent f is basically reliable.