CAJ | 학술논문

针对BP神经网络搜索速度慢、易陷入局部极值的缺陷，采用PSO算法优化BP神经网络后建立各影响因素与部分碳化区长度的关系模型。将改进后的模型进行实验仿真训练并应用于某混凝土大桥部分位置的碳化深度预测中，仿真应用结果表明，网络输出值和期望值很好吻合，收敛速度更快。所以该模型能够对混凝土部分碳化区长度进行预测，为混凝土结构耐久性设计、评估和寿命预测提供科学指导。
침대BP신경망락수색속도만、역함입국부겁치적결함，채용PSO산법우화BP신경망락후건립각영향인소여부분탄화구장도적관계모형。장개진후적모형진행실험방진훈련병응용우모혼응토대교부분위치적탄화심도예측중，방진응용결과표명，망락수출치화기망치흔호문합，수렴속도경쾌。소이해모형능구대혼응토부분탄화구장도진행예측，위혼응토결구내구성설계、평고화수명예측제공과학지도。
For the defects on search speed slow and ease into the local minima of the BP neural network, the model about each factor and the length of the partial carbonation zone is established, which uses the PSO algorithm to optimize the BP neural network. Based on the experimental simulation training, the improved model is applied to the part location of a con-crete bridge on carbonation depth prediction and the simulation application results show that the network, having faster convergence rate, output values and expectations is a good match. So the model can predict the length of the partial car-bonation zone of the reinforced concrete, which offers scientific guidance to durability design, assessment of concrete structures and life prediction.