Research On Residual Life Prediction Model Based On Bridge Remote Monitoring Information

During service of bridge, coupling effects of unfavorable factors such as environmental erosion, material aging, long-tem effect of load, fatigue effect and mutagenic effect lead to damage accumulation and resistance attenuation in the structure and system inevitably, and residual service life of bridge endure severe test. With the development of bridge health monitoring and improvement in accuracy of test information, at the same time, the time-varying reliability theory is consummated continuously; the above improve reliability of prediction and evaluation mostly. Because existing prediction method for structure service life can't make reliable prediction accurately and systematically, this task form studying on real-time monitoring information of bridge and based on time series theory and reliability theory establish mathematical statistical model between structural resistance and load effect, in order to predicting residual service life of bridge's whole structure.Firstly, advantage and disadvantage of usual prediction method for structure service life are compared, the characteristic and advantage of life prediction based on monitoring information is analyzed; the three measures of confirming bridge's initial state are proposed using code, finite element theory and monitoring means. The factors of affecting prediction for bridge reliable residual life are consist of structural resistance and load effect, so the topic of this paper is composed form these. Through the materials of studying on prediction measures of the structure service life overseas are collected , kinds of influencing factors of inducing bridge's resistance decay are summarized, and influence mechanisms are analyzed deeply. Using the basic theories of material mechanics and structure dynamics, characteristic factors of reflecting structure resistance decay are deduced. The theory time-frequency analysis reveals the interdependence between monitoring information and these characteristic factors. Based on the measure of mathematics and the identify of these characteristic factors, the way that these characteristic factors reflect structure resistance is lodged. Analysis character of wavelet multi-scale and mathematical fitting method provides thoughts for question of live load effects separation in monitoring information, the above offer a correction method for existing load method.Finally, using time series theory, structure resistance based on monitoring information and time-varying mode of load effect are established, the processes establish frame diagram of life prediction of the reliability structure.Based on monitoring information of actual bridges, the dynamic trend of the structure resistance and the effect of live load are analyzed, the function expression which is ARMA prediction of deflection is proposed, and accuracy of life prediction model is summarized and verified.Various knowledge such as information, mathematics and mechanics are crossed in this model. The thought have a wide range of engineering applications, and significant social and economic benefits.