Common anomaly recognition methods are easy to misjudge and miss outliers for the online monitoring data. This is a bottleneck problem that needs to be overcome in dam safety management moving toward informatization. Based on the data of nine hydropower stations along Dadu River Basin, this paper analyzed existing problems of the common anomaly identification method and an algorithm was proposed based on improved M-robust regression recognition. In this algorithm, the AR factor was introduced to avoid the defect that the traditional model cannot simulate random variables. The extreme value method and robust estimation were utilized to avoid the leverage effect. The model collapse caused by maximum measured value was avoided through improving the residual calculation model of M-robust and optimizing the weight distribution function. The maximum of the three values, residual quartile difference, discrete quartile difference, and measurement accuracy, was used as an anomaly recognition criterion to improve the evaluation criteria. The algorithm compiled was used in the Dadu River Company since 2017. The statistics showed that for the 150,000 measured values per day, the evaluation time could be within 15 min, the missed judgment rate was 0%, and the misjudgment rate was less than 2%. The proposed algorithm achieved a great improvement and can meet the needs of online outlier recognition in dam safety management.

Rockfill dams are the most competitive type of dam in complex geological environments. Identifying the failure modes and causes in high dams over 100 m is critical for better guiding high dam designs and implementing safety prevention and control measures. To this end, this paper investigated numerous cases of earth–rock dam breaches and failure modes in rockfill dams globally, with a particular focus on dams over 100 m in height, encompassing all such dams in China. The study categorized dam failure modes based on whether the dams were built before or after 1980. It also examined the causes of dam failures in terms of dam height, foundation characteristics and thickness, and failure time. Additionally, the paper analyzed a rockfill dam in China, with a height of 136 m and over ten years of operation, as a case study. We analyzed the spatial and temporal characteristics and causes of failures, such as dam crest cracking, high-level seepage, and gallery cracking, using the design situation, monitoring data, and numerical simulation. The paper also addressed issues related to dam design and foundation treatment, providing recommendations for improvement. The study indicated that the overall risk of total failure for dams over 100 m is already low. However, longitudinal cracks on the dam crest, core wall seepage, hydraulic splitting, and seepage damage to the dam foundation are primary issues in the current high core wall rockfill dams. These issues are mainly caused by uneven structural deformation of the dam and its foundation. A reasonable design of rockfill materials and foundations can mitigate these failures.

Hydropower stations, as a green energy source, have grown rapidly in China, and dam safety is critical to ensuring power generation efficiency. There is currently no safety monitoring and management platform that combines dam structure, geology, hydrology and other data. To this end, a dam safety monitoring information management system is built based on a GIS platform, with ArcGIS Engine serving as the GIS class library and the NET Framework serving as the programming model for program implementation. The system connects the GIS spatial database to the monitoring database, combines spatial analysis and data mining technology, and realizes functions such as 3D visual structures display, spatial analysis and evaluation of monitoring data, providing a more scientific and intelligent method for dam safety management. After verification in actual projects, the system can reasonably assess dam safety using measured data and projects, as well as make decision support recommendations.

High dams are constructed in mountain valley areas with extremely complex geomorphologies, geologies, and operating environments. The dam body and foundation are subjected to large hydraulic loadings and high stress levels, where many factors affect the structural deformation of the dam, making dam safety monitoring and risk assessment difficult. Since dam deformation is critical for evaluating safe operation, a novel dam deformation monitoring model based on the component separation method is proposed, which can provide scientific significance and high application value for operational safety control. The deformation monitoring data include deformation caused by the water level, the temperature, and the time effect. The homologous monitoring sequence points are selected, and a safety monitoring model is proposed based on the physical causes and the independent component signal decomposition technology. First, the aging component can be extracted by using the correlation consistency of a specific measurement point, where the environment is almost the same at the same time of year, such as the temperature and the water level. Then, fast independent component analysis algorithm is applied to further extract the dam deformation induced by the water level and the temperature via correlation analysis. Finally, a deformation monitoring model based on the component separation can be successfully constructed by combining models of the three components. The proposed dam deformation model can overcome problems of large covariance of deformation component, large collinear interference of periodic component separation, and unstable separation. It is further applied to a high face rockfill dam, and the results shows that the multiple correlation coefficient is greater than 0.995 and the mean absolute percentage error is less than 1.37%, demonstrating the high accuracy and stable separation of the components produced by using this model.