综述了BP神经网络在混凝土抗压强度及其他力学性能、耐久性、混凝土结构构件相关性能预测中的应用.综述表明,模型误差在工程可控范围内,模型泛化能力较强,利用BP神经网络对混凝土性能开展预测是可行的.
This paper reviews application of BP neural network in prediction of concrete performance,including com-pressive strength and other mechanical properties,durability and properties of concrete structural members.Error of the models is basically within a controllable scope of engineering practice.In this context,BP neural network has ex-hibited a sufficient generalization ability to implement a prediction of concrete performance.