论文部分内容阅读
针对目前广泛使用的模糊系统和神经网络预测方法在地下结构围岩变形预测中的缺陷,提出一种精确在线支持向量机(AOSVR)并将其应用到水电站地下厂房开挖过程中项拱围岩的变形预测。通过与其他预测方法的比较,可以发现精确在线支持向量机有很强的学习能力和很高的预测精度。
Aiming at the shortcomings of fuzzy systems and neural network prediction methods currently widely used in the deformation prediction of surrounding rock mass in underground structures, an accurate online support vector machine (AOSVR) is proposed and applied to the surrounding rock of the arch in the underground powerhouse of hydropower station Deformation prediction. Comparing with other prediction methods, we can find that the precise online support vector machine has strong learning ability and high prediction accuracy.