A method for automatic identification of dam anomaly monitoring data

Abstract

The invention discloses a dam exception monitoring data automatic identification method. The method comprises the steps of constructing a track matrix of an original monitoring data sequence, and performing singular value decomposition on the track matrix to obtain feature sets; arranging the feature sets according to eigenvalues from high to low, and selecting the first multiple feature sets with an accumulated contribution rate greater than or equal to 85% as main feature sets; calculating a basic matrix corresponding to the main feature sets, and performing diagonal averaging on the basic matrix to obtain main components of the data sequence; accumulating the main components to obtain a reconstructed data sequence; performing subtraction operation on the reconstructed sequence and the original data sequence to obtain a residual error sequence, and calculating out a standard deviation of the residual error sequence; and judging whether a measured value is an abnormal value or not by a Pauta criterion according to a residual error of the measured value. According to the method, main features of the monitoring data sequence can be automatically extracted, so that a mathematic model is prevented from being built by manpower, the consistency and accuracy of judgment can be ensured, and manpower resource input is reduced; and when environmental variables such as water level, air temperature and the like are deficient, monitoring data still can be judged.

Description

technical field [0001] The invention relates to the field of dam monitoring, in particular to an automatic identification method for abnormal monitoring data of a dam. Background technique [0002] my country has built more than 200 dams over 100 meters in total, including more than 40 high dams over 150 meters. Monitoring and monitoring these dams is an important means to ensure their safe operation. The "Regulations on Supervision and Management of Hydropower Dam Operation Safety" promulgated by the Energy Bureau in 2015 further requires that "for dams with a dam height of Record) registered dam operation safety for remote online technical supervision". Due to the complex working conditions, high dams generally have far more monitoring points than general projects; and when managing dams, the number of measuring points involved is even greater. [0003] An important task of online monitoring is to identify abnormal operating conditions, and its basis is to identify abno...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem of excessive dependence on manual operation and environmental quantity in the dam abnormal monitoring data identification technology based on statistical models, the present invention provides a method for automatically identifying abnormal monitoring data, which aims to improve abnormal detection in dam safety monitoring work. Accuracy, efficiency and robustness of value recognition, and reduce human resource investment

Images