| More than 98,000 reservoirs and dams have been built in China,of which the earth dams account for about 95%.A large number of high core rockfill dams have been put into construction and operation with the further development of national hydropower energy,such as the Pubugou(dam height 186 m),the Nuozhadu(dam height 261.5 m),the Lianghekou(dam height 295 m),the Shuangjiangkou(dam height 312 m),etc.For these large-scale basin control projects,the dam height is high,the reservoir capacity is large,and they are located in a complex geological environment with the characteristics of large loads in the dam body and foundation,high-stress level,and complex failure formation mechanism and risk time-varying laws,thereby leading to the difficulty in monitoring the safety risk intelligently.Therefore,solving the difficulty in the real-time perception of dam status information,accurate identification of abnormal data,and dynamic assessment of safety risk is the key scientific and technological problems to realize the intelligent identification and early warning control of dam operation safety risk.At present,many scholars in the world have carried out much research in the field of dam safety risk intelligent analysis and the dam safety management system,and rich achievements have been made.However,there are still some theories and techniques that have not been solved systematically in the aspects of the intelligent acquisition of monitoring data,online identification of abnormal data,and dynamic assessment of safety risk.Therefore,it is of great theoretical significance and application value to making a study on the key technologies of operation safety risk intelligent monitoring of high core rockfill dams.Given the key problems of information perception,data identification,and risk assessment involved in the operation safety risk intelligent monitoring of high core rockfill dams,an in-depth study of the key technologies such as the intelligent monitoring of dam external deformation,the online identification of abnormal data,and the dynamic assessment of dam crest cracking risk and so on has been made in the presented paper based on the theoretical analysis,field test,numerical simulation,system research and development,and engineering practice method.Some innovation conclusions are as follows:(1)A baseline calibration and meteorological fusion correction model is constructed,a datum point stability check method for TPS-GNSS data fusion is proposed,and a TPS-GNSS-based integrated intelligent station of dam external deformation is developed,forming an intelligent monitoring technology of dam external deformation based on TPS-GNSS.In view that the deformation monitoring of dam and slope is greatly affected by the mountain environment,and the monitoring accuracy and intelligent degree are low,a baseline calibration and meteorological fusion correction model considering the representative error of meteorological observations and the uncertainty of atmospheric vertical refraction coefficient by combining the TPS and GNSS technologies is established,a real-time datum point stability calibration method for TPS-GNSS data fusion based on the least square adjustment test method and the average gap method is proposed.And an integrated intelligent monitoring station for dam external deformation is developed,which is based on TPS and supplemented by GNSS and has the functions of large field angle,environment self-adaptive,thereby forming an intelligent monitoring technology of dam external deformation based on TPS-GNSS and effectively solving the problems of small monitoring range,low monitoring accuracy under extreme weather conditions,poor timeliness of spatial data analysis and low degree of intelligence in the external deformation automatic monitoring system of dams and high slopes.The application of the PB dam shows that the station has the advantages of reliable results,high efficiency,stable operation,good performance,and remarkable economic benefits.It can provide effective technical support for the external deformation monitoring of high dams in the mountainous environment and has important engineering application and popularization value.(2)A complete set of online identification technology of three-stage progressive data anomaly in dam safety monitoring is established.Aiming at the problems such as the multiple types of dam safety monitoring data sequences,low accuracy of anomaly identification and heavy workload,an in-depth study of the dam safety monitoring abnormal date of the causes,identification method,identification model and process of and so on.A threshold criterion based on M-estimator and confidence interval radius(MZ criterion),and an MZ criterion-based online robust identification model for data anomaly are established,and a three-stage progressive deduction simulation algorithm that integrates the abnormal mutation identification of single-point time series simulation,measurement error reduction by using the remote retest space-time characteristic analysis and the mutation identification of environmental response and its technical process is proposed,thus realizing the online classification and identification of abnormal causes of dam monitoring data,solving the missing judgment and misjudgment problems in the monitoring data sequence,and then forming the complete set of technologies for online identification of dam safety monitoring data.The application of the comprehensive management platform of reservoir dam safety information in the basin where the PB dam is located shows that the technology is reliable,and has high identification accuracy and efficiency.The misjudgment and omission rate of abnormal identification of monitoring data has been reduced to be less than 2%,the online identification is achieved and the non-structural abnormal changes such as the accidental error,instrument failure,and environmental response are eliminated,which provides reliable data support for dam safety risk monitoring and has important theoretical significance and engineering application value.(3)The dynamic assessment and prediction model of crest crack risk in high core rockfill dams are established,and an operation safety risk intelligent monitoring system for high core rockfill dams is developed.Aiming at the difficulty of risk assessment and early warning of dam crest cracking,and the low intelligent degree of dam operation safety management and so on in the high core rockfill dams during its service life,the performance function agent model of dam crest cracking and the real-time assessment model for of dam crest cracking based on the dam deformation monitoring are constructed considering the criteria fuzziness of dam crest cracking,the randomness of dam and foundation material parameters and the nonlinear coupling relationship between cracking risk and dam deformation.Taking the duration time of abnormal state in real-time risk assessment as the random variable,a dynamic risk assessment and prediction model for dam crest cracking is established,and the dynamic identification technology of risk model parameters driven by subjective and objective information based on the Bayesian theory is proposed.An operation safety risk intelligent monitoring system integrating the monitoring data acquisition of external deformation,online identification of data anomaly,and dynamic assessment and early warning of dam crest cracking for the PB high core rockfill dam is developed,thereby realizing the classification and identification of dam monitoring abnormal data,dynamic analysis of the time-varying probability of dam crest cracking and early warning of cracking risk,etc.Engineering application shows that the model has high parameter identification,the simulation results accord with the engineering practice,and the operating system is stable and reliable,which provides important technical support for intelligent monitoring of operation safety of high core rockfill dams. |
PDF Full Text Request