摘要
针对四川省映秀湾电站地下厂房顶拱混凝土裂缝特征 ,结合神经网络的特点 ,建立了裂缝开度的BP网络模型。经实例模拟预测分析表明 ,模型预测精度较好 ,方法可行 ,应用BP人工神经网络预测地下厂房顶拱裂缝观测值 ,避免了裂缝值与其影响因素之间难以用数学式精确表达的问题 ,同时也克服了由于各因素的不确定性所造成的影响。根据裂缝多年的发展趋势与BP模型的预测结果 。
This paper focuses on the characterstics of cracks of Ying xiuwan underground powerhouse arch crowns in Sichuan Province and establishes crack opening BP model.The application of practical prediction shows that the precision of model is satisfying.The method avoids the difficulty of expressing the relation between the crack opening value and its influence factors.Meanwhile the influence of uncertainty of every factors is also avoided.According to the trend of crack and the precision of BP model,reinforcement measures must be takern quickly.
出处
《四川大学学报（工程科学版）》
EI
CAS
CSCD
2002年第2期21-23,共3页
Journal of Sichuan University (Engineering Science Edition)
