Statistical Inference Methods Of Dead Loads And Material Strength On Existing Structures

The reliability of existing structure is an important embranchment of the reliable theory. The methods of appraising the existing structural reliability that are used in our country generally belong to applied methods. But the existing structures have already converted into realistic substantiality, and they have much characteristic different from the structures in design, so there are many illogical aspects in appraising the existing structures based on the design criterion. Compared to applied methods, probability methods are more reasonable, because of considering the characteristic of existing structures.The paper aimed at the characteristic of existing structures, which is carried out small samples and bayes statistical inference methods of dead loads and material strength on existing structures, and will impel the methods of appraising existing structural reliability transform from applied methods to probability methods. Three instances are discussed in this paper: the unknown variance and variant coefficient, the known variance and the unknown variant coefficient. According to matrix information and road-test dates, the small samples and bayes statistical inference methods are used for studying the standard values of dead loads and material strength on existing structures in these instances. The paper, particularly discussed the basic idea and theory of bayes methods, and brings forward separately the bayes statistical inference methods with small samples of the standard value of material strength on the unknown variance and known variance. It also contrasts the traditional methods with bayes methods, points out the distinction of inference results, and that recommends relatively advantageous inference methods.Now there have not particular study about the inference of existing structural dead loads and material strength when the parameter is unknown. The disquisitive progeny of this paper not only accords with the reference of existing structural reliability, butalso makes for estimating the existing structural reliability reasonably. So the disquisitive progeny of this paper may provide reference for the actual standard used in our country, and it will also impel the methods of appraising the existing structure develop from applied methods to probability methods.