According to the characteristics of self-anchored suspension bridge, a new method to detect damage is introduced in this paper.It works in two stages.First, a BP neural network model is built to predict damaged position. Next, based on the characteristics of genetic algorithm and simulated annealing algorithm, a new approach, genetic-simulated annealing algorithm, is put forward to identify damage extent of detected positions. Compared with the traditional genetic algorithm, the global convergence effect of this algorithm is enhanced by using of the Metropolis acceptance rule of the simulated annealing algorithm in the searching process.

Parameter changes of a bridge in construction process must have an effect on its stress and alignment.And different parameters ,in the same bridge ,affect state of bridge to a different degree. While we have no need to consider every parameter in practical construction process, just paying attention to some sensitive parameters with a great effect on bridge structure instead .So a reseach of parameter sensibility of bridge construction is necessary to find out sensitive ones of all to limit their great effect .In this paper, based on JinZhou cable-stayed bridge without backstays,we use Midas civil to analyze the degree of impact of each parameter on stress and alignment of main girde ,which is depended on to put forward the main sensitive parameters of the bridge in construction process and a few of advice about its construction technology.