论文部分内容阅读
大坝升船机顶部厂房与底部塔柱结构之间的侧移刚度突变造成了顶部厂房强烈的地震鞭梢效应。针对此问题,本文进行了屋盖MR智能隔震系统对大坝升船机地震鞭梢效应智能控制的振动台试验研究。屋盖MR智能隔震系统由隔震垫、磁流变(MR)阻尼器及控制系统组成,安装在厂房的柱子顶部。首先采用修正的Bingham模型得到数据来训练神经网络以反映MR阻尼器的逆向动力特性,然后与模糊算法结合通过dSPACE实时控制系统得到阻尼器的最优控制信号。试验结果表明:模糊-神经控制方法能有效控制屋盖MR智能隔震系统,使得大坝升船机顶部厂房的鞭梢效应明显减小,排架柱顶位移减小了50%,隔震层相对位移减小了57%。
The sudden lateral displacement stiffness between the top of the shiplift and the tower structure at the bottom of the ship caused a strong earthquake whipping effect on the roof of the plant. In response to this problem, this paper carried out a shaking table test study on intelligent control of seismic whipping effect of dam lift ship with intelligent roof MR system. Roof MR smart isolation system consists of a shock pad, MR damper and control system installed on the top of the plant’s pillars. Firstly, the modified Bingham model is used to train the neural network to reflect the inverse dynamic characteristics of the MR damper. Then, the optimal control signal of the damper is obtained through dSPACE real-time control system combined with the fuzzy algorithm. The experimental results show that the fuzzy-neural control method can effectively control the MR intelligent isolation system of the roof and make the whiplash effect on the top of the shiplift significantly reduced, and the displacement of the top of the berth reduced by 50% Relative displacement decreased by 57%.