| Transportation is the artery of economic development. It is inseparable that the economies of the area gets in toughly touch with highway communication. But because of the long term open to the traffic, natural environment affections, concrete structure aging, bridges' health status degraded constantly, which seriously influences the traffic safety. It is particularly important to evaluate the health status of bridges.The traditional evaluating bases on finite element model and healthily tests measure materials. Finite element model needs assumption or simplification to the border of the structures. In this aspect, professional knowledge is needed, which make these methods difficult adopted by general administrative staffs.This paper summarizes research results in the past and adopts BP neural network method to evaluate the health status of bridges. BP neural network method is based on a large number of statistics dates from design,test,maintenance and so on.BP neural network is a kind of relation of shining upon non-linearly, which can draw inherent relation from data-in train. It needn't rigorous structural theory and relevant advanced professional knowledge. It is easy to be accepted and applied.In this paper, the writer has done the following works:â‘ Improved the training parameter of the BP neural network, making it more suitable for bridge health status evaluating.â‘¡Written the bridge health database software (BHMS).The software offers the dates needed for the evaluation in the future. At the same time, it can make the administrator more concrete at time course change of the bridge health status.â‘¢ Contrasting BP neural network method with other methods to pro(?)e BP n(?)ural network method is feasible. |
