Deformation monitoring is the main program in the area of dam safety. Because statistical model is simple and intuitive, it is widely used in dam safety monitoring. However, in dam's displacement statistic model, there is a high degree of linear relationship between influence factors. Due to the influence of multicollinearity, models calculated with traditional methods are not accurate and stable. Besides, because of dam integrity, each part of dam is interrelated and interactive. Currently, single point or multipoints displacement monitoring models cannot accurately reflect the actual dam running state. In this paper, the theory of panel data is introduced to dam deformation analysis. Panel data contain time series data and cross section data, which is able to solve serious multicollinearity problem of traditional regression method. Moreover, all measuring points are classified into several groups according to their similar deformation law. Based on the random-coefficient model of panel data, potential relationship between different measuring points is built. Take 1 hydropower station, for example, to examine that random-coefficient model is able to improve the modeling situation that estimators are not significant and simultaneously provide a stable model, which explores a new approach for the research of dam displacement monitoring. [4] used statistical regression method to research dam prototype observation data. He set deformation measured value as explained variable and set appropriate explanation variables at the same time. The unbiased efficient estimator of each explanation variable was obtained through the traditional least squares method (ordinary least squares, OLS).After that, influence factors of deformation-measured values were analyzed quantitatively. On the basis of Chen JY's study,
Displacement data modelling is of great importance for the safety control of concrete dams. The commonly used artificial intelligence method modelled the displacement data at each monitoring point individually, i.e., the data correlations between the monitoring points are overlooked, which leads to the over-fitting problem and the limitations in the generalization of model. A novel model combines Gaussian mixture model and Iterative self-organizing data analysing (ISODATA-GMM) clustering and the random coefficient method is proposed in this article, which takes the temporal-spatial correlation among the monitoring points into account. By taking the temporal-spatial correlation among the monitoring points into account and building models for all the points simultaneously, the random coefficient model improves the generalization ability of the model through reducing the number of free model variables. Since the random coefficient model supposed the data follows normal distributions, we use an ISODATA-GMM clustering algorithm to classify the measuring points into several groups according to its temporal and spatial characteristics, so that each group follows one distribution. Our model has the advantage of having a stronger generalization ability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.