The Uplift Pressure Monitoring Data And The Forecasting System Of The Concrete Dam

Based on the uplift pressure monitoring data analysis and comprehensive evaluation,it is very singificant to analyze the monitoring data of uplift pressure for checking the stabilization,monitoring the seepage feature,learning about the work efficiency of foundation curtain and drainage system,understanding a sort of load condition that affects the dam distortion and stress.The paper discusses the necessity of the dam safety monitoring work and analyzes the main factors influencing the uplift pressure.According to the characteristics of dam monitoring data,relative merits of the forecast models and their applied range adopted generally at present are analyzed.Establishing gray models(the equidistant sequence gray model,and the optimized gray model)and the mending BP neural network models(the gradient decent algorithm having momentum gene and rebound BP algorithm)for forecasting uplift pressure of dam,The software of the forecasting system is exploited with Delphi7.0,Matlab6.1,ActiveX technology supplied by Matlab to obtain the interface between Matlab and Delphi,using ADO objects to connect Access database.Studying the LONGYANGXIA dam,the paper dissertates the composing and the working principles of monitoring system,analysises of riverbed dam section in 2443m height on 98 effective observation point uplift pressure monitoring data,determine the curtain effect of significant position is obtained,and the left side and the right side of riverbed uplift pressure of basic patterns are consistent with increased year by year of the upstream water level variation,high water level of river bed curtain has a greater influence,existence the emergence of new possibility of cracks,and the application of uplift pressure of concrete dams prediction system on measuring point D8P21-17 uplift pressure prediction.The practice suggested that the results of forecasting are satisfied with gray models and BP neural network models.Sequentially,it is proved that the software is valid and practicable.