摘要
为了提高模型预测的精度,根据灰色理论,建立超长桩承载力的非等步长灰色微分方程,得到了微分方程的精确解。基于优化理论,建立误差目标函数。利用最小二乘法,直接求得微分方程的系统参数,得到灰色优化模型。以此模型为基础划分n个状态,应用Markov原理,通过计算状态的一步转移概率矩阵,建立了灰色优化-马尔柯夫预测模型。使用该模型预测超长桩的承载力,预测值和试验值吻合较好,与GM(1,1)模型、灰色优化模型相比,预测精度进一步提高,为超长桩承载力的预测提供了新的方法。
Based on gray theory, an unequal step length gray differential equation was presented to describe bearing capacity of ovedength piles, and obtain its accurate solution. Error objective function was created based on optimization theory, further used least squares method to get the system parameters of the differential equation. In order to improve the accuracy of model, according to Markov theory, dividing the test value of gray optimizing model into n kinds of states based on predicating fitting curve, and getting one-step transfer matrix to forecast beating capacity of ovedength piles. The predication results show that the gray optimizing -Markov model based on the load-settlement curve can improve the precision of prediction compared with GM（1, 1） model and gray optimizing model. A new way is provided for the predication of bearing capacity of overlength piles.
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2006年第3期423-427,434,共6页
Rock and Soil Mechanics
基金
温州市科技计划项目(No.S2004A007)
