Simulation of Local Outlier Warning Method for Dam Risks in Dangerous Reservoirs
The anomalies in the monitoring data of reservoir dams may reveal localized risks,and if they are not accurately identified and responded to,the potential risks will be neglected,affecting the warning accuracy.In order to ensure that the risk of dams at risk can be accurately early warned,a method of early warning of the risk of dams at risk based on the localized coefficients of the outliers is proposed.The topographic and hydrological parameters of the dam are extracted from the digital elevation model,and the risk assessment is carried out by combining different pa-rameters.According to the results of the risk assessment,the area of the dam is divided into different risk levels to re-alize the division of the risk area of the dam at risk.The partial least squares method is used to analyze the correlation between topographic parameters and hydrological parameters,obtain the orthogonal feature space,introduce the local outlier coefficients to judge the data anomalies in the monitoring data in the feature space,and realize the risk warning of dams in reservoirs at risk according to the results of the anomalous data judgment.The experimental results show that the proposed method can effectively improve the accuracy of the risk warning results of dams at risk,which is of great significance to ensure the safety of dams.
Local outlier coefficientDangerous reservoir damsRisk warningDigital elevation modelingPartial least squares
万方数据
维普
