Applied Research Of Wavelet Theory In Dam Deformation Monitoring Data Analysis

The overall behavior of dam mainly reflected in the deformation, seepage, stress-strain and other changes, of which changes in the situation deformation is particularly intuitive and reliable, the deformation behavior comprehensively reflect the impact of various factors of its work behavior, and the change is an important basis for the evaluation of dam safety. There are some limitations in deformation analysis models algorithms, application effects and application ranges of the existing domestic and international. Therefore, it is an important task that how to use new theories, new methods to overcome effectively the shortcomings of traditional modeling methods to resolve the key issues of modeling technology in dam safety monitoring. This test exactly proceeds from this angle. Aiming at some problems of dam safety monitoring data analysis and modeling, introduce wavelet theory researching on the wavelet analysis applications in dam monitoring data analysis of deformation monitoring. The research contents and main results acquired can be summarized as follows:(1) Analysised variously to dam deformation monitoring data using the wavelet theory. Analysised the wavelet analysis application principles and ideas systematically in signal anomaly detection and handling, de-noising and trends extraction, and then developed a wavelet analysis program based on MATLAB. Examples showed that wavelet analysis is suitable for post-processing of dam safety monitoring data, it was less time-consuming and high accuracy than other methods.(2) Established the dam deformation monitoring wavelet neural network prediction model based on adaptive adjustment of learning rate optimization algorithm.Through improving traditional Wavelet Neural Network (WNN) algorithm, it achieved truly the organic integration and penetration of wavelet theory and neural network, and wrote the corresponding modeling and analysis procedures by taking advantages of MATLAB software powerful numerical calculation and simulation analysis function. (3) The built dam defomation monitoring based wavelete neural network model was applied to engineering example research. The trained wavelet neural network model was applied to the deformation of a dam monitoring fitting and forecasting, and it also has been made a comparative analysis with conventional BP neural network prediction results. The results showed that the wavelet neural network model based on adaptive adjustment of learning rate optimization algorithm had high precision than the BP network model, and it has good practical value.