基于RBF神经网络的变电站混凝土立柱抗压强度评定

Assessment of Compressive Strength of Substation Concrete Column Based on RBF Neural Network

变电站混凝土立柱抗压强度的评定是判断变电站混凝土结构损伤程度、剩余承载力的重要依据。设计了一个RBF神经网络模型,将其应用于超声回弹综合法评定变电站混凝土立柱抗压强度,给出了用超声回弹法进行混凝土强度测试的方法。经试验测试和仿真分析表明,所提出的RBF神经网络比传统的回归计算方法具有更高的评估精度。

Assessment of compressive strength of substation concrete column is an important foundation of damage degree and bearing capacity of construction. An RBF neural network model （RBF-NN）is applied to assessing compressive strength of concrete by ultrasonic and rebound combined method. An experimental method is given for compressive strength of concrete test by ultrasonic and rebound combined method. It is proved that RBF- NN model has higher evaluation precision than that of regression calculation by experimental test and emulation analysis.