结构基于RBF神经网络的变结构控制研究

New variable structure control for buildings using RBF neural network

针对固定增益变结构控制器应用于结构振动控制工程中可能引起控制系统过大的抖振,采用了一种基于RBF神经网络调节控制切换增益的变结构控制方法对地震作用下建筑结构的振动控制问题进行研究。根据二次型最优控制方法设计了切换面;基于切换面存在的可达性条件推导出控制律的表达式。以一个三层框架结构模型为例来验证所提变结构控制方法的有效性,算例分析结果表明,所提控制方法能够有效地减小结构的地震峰值响应,且达到了削弱控制系统抖振的目的。

As to the conventional variable structure control methods for reducing the seismic responses of building structures, the shortcoming is that it is very difficult to avoid the excessive chattering effect. Therefore, based on the advantage of RBF neural network control method, we propose a new adaptive variable structure control approach that can not only reduce the seismic responses but also avoid the undue chattering effect. We design the switch sur- face according to the linear quadratic optimal control algorithm and deduce the control law according to the exist- ence and reaching condition of the switch surface. Finally, we take a three-storey building＇ s model as a numerical example to verify the effectiveness of the controller. The ground accelerations recorded in two different earthquake events are used to evaluate the effectiveness of the control algorithm for various disturbances. The simulation results show preliminarily that our new adaptive variable structure control method is quite effective. It can not only reduce the peak-response of the ground motion, but also keep the chattering effect sufficiently low so as to ensure the sta- bility of the system.