论文部分内容阅读
为了准确模拟大跨空间结构的风荷载,提出了改进的线性滤波自回归模型(Auto-Regressive model)法。该方法基于传统线性滤波自回归模型,通过高斯-赛德尔(Gauss-Seidel)迭代法求解大跨空间结构风荷载模拟中的回归系数矩阵,进而对大跨空间结构的风荷载进行模拟。结果表明:改进的线性滤波AR模型法解决了自由度过大导致的回归系数矩阵不正定的问题;能够有效地模拟具有空间相关性、时间相关性的节点脉动风速时程,模拟精度、速度和计算稳定性均满足实际工程应用要求。
In order to accurately simulate the wind load of a long-span space structure, an improved linear regression model (Auto-Regressive model) is proposed. The method is based on the traditional linear filter autoregressive model and the Gaussian-Seidel iteration method is used to solve the regression coefficient matrix in the wind load simulation of long-span space structures. Then the wind load of the long-span space structure is simulated. The results show that the improved linear filter AR model solves the problem of the indeterminacy of the regression coefficient matrix caused by too large degree of freedom. It can effectively simulate the time-dependent, time-dependent, Computational stability to meet the actual project application requirements.