基于多维AR模型的桥梁随机风场模拟

Simulation of bridge stochastic wind field using multi-variate Auto-Regressive model

基于自然风特性,通过考虑结构节点间风速时程的空间相关性,采用多维AR模型模拟主梁和桥墩节点随机脉动风速时程,利用FPE准则和AIC准则确定模型阶数,并对模拟过程中的自回归顺序、功率谱密度等问题进行研究。对兰新二线铁路白杨河大桥采用多维AR模型模拟各节点的脉动风速时程,结果表明:当AR模型阶数为4时,模拟功率谱与目标功率谱吻合较好;当自回归顺序颠倒时,模拟功率谱明显偏离目标功率谱。

Based on the natural wind properties and the correlativity of nodal wind speed time history, a multi-variate auto-regressive （MVAR） model was used to simulate the time history of fluctuating wind for the main beam and bridge pier, and FPE and AIC rules were employed to determine the order of MVAR model, then related problems such as auto-regressive sequence and power spectrum density were discussed. Taking the Baiyang River Bridge on the Second Lanzhou-Urumqi railway as an example, the wind speed time histories were simulated with the MVAR model. The results show that the simulated power spectrum density is consistent with the target one when the model order is equal to 4, however, it shows obvious deviation when the auto regressive sequence is reversed.