Timely damage detection and identification of a large number of existing building structures has always been necessary. Bayesian updating based methods can solve the problem of uncertainty and unsuitability during measurement with a small amount of measured data. In this paper, the Bayesian update of the target member is simulated under the real detection condition, and the sequence update of the target member is tried with the increase of the measured data. This paper concludes that detecting and identifying structural damage based on Bayesian updating has high detection efficiency and accuracy, and has practical significance in the field of engineering structure inspection. The method can continuously update the a priori probability to obtain the posterior probability under small sample and non-complete information conditions by adding events to the original a priori probability to infer the structure’s damage state. By combining prior knowledge and real-time data, the Bayesian method can provide reliable damage assessment results in uncertain environments, and improve the accuracy and decision support capability of structural health monitoring. And as the measured data is added and the number of updates increases, the updated results will become more accurate. In the future, a part of the truss can also be used as a Bayesian updating unit, and the overall damage detection and identification of the truss can be completed more efficiently through the simultaneous Bayesian updating of multiple rods.