现行混凝土结构设计规范中钢筋混凝土柱的受剪承载力公式为半经验半理论计算公式,由于试验数据的有限性和钢筋混凝土材料离散性较大的本质特性,规范建议公式缺乏明确的理论模型。该文以贝叶斯动态信息更新思路,根据主观经验信息选定现行规范中钢筋混凝土柱受剪承载力计算公式作为贝叶斯先验模型,通过已完成试验研究的柱受剪承载力数据,应用贝叶斯方法综合这两类信息进行推断,建立钢筋混凝土柱在反复荷载作用下受剪承载力的概率模型,对未知参数进行估计,修正先验模型并建立后验模型,以达到对先验模型的更新。通过先验模型计算值、后验模型计算值及试验值相互间的对比分析可知:贝叶斯方法继承了先验信息的完备性和大量试验数据的准确性,能够更准确地预测钢筋混凝土柱的受剪承载力。 The formula of shear capacity of RC columns in current design code of reinforced concrete is semi-empirical semi-theoretical formula. Due to the limited experimental data and the inherent nature of reinforced concrete material with large discreteness, the specification formula is lack of a clear theoretical model. Based on the Bayesian dynamic information renewal method, according to the subjective experience information, the formula of shear strength of RC columns is selected as Bayesian a priori model. Based on the data of column shear capacity, Applying the Bayesian method to infer these two kinds of information, the probabilistic model of reinforced concrete columns subjected to repeated loading under shear load is established, the unknown parameters are estimated, the a priori model is modified and the posterior model is established to achieve the first Update the model. By comparing the calculated values ​​of the a priori models with the calculated values ​​of the posterior model and the experimental values, we can see that the Bayesian method inherits the completeness of the prior information and the accuracy of a large number of experimental data to predict the reinforced concrete columns more accurately Shear capacity.