结构基于独立模态空间控制的神经网络控制

THE NEURAL NETWORK CONTROL BASED ON INDEPENDENT MODAL SPACE CONTROL OF STRUCTURE

在对柔性结构振动的主动控制中,独立模态空间控制是一种适合于柔性结构振动控制的方法,但如果控制器选择不好就会带来观测溢出和控制溢出.神经网络控制则具有较好的泛化作用和鲁棒性,对抑制观测溢出和控制溢出具有较好的效果.本文在独立模态空间控制的基础上设计了一种神经网络控制,并对柔性梁及板进行了控制的数值仿真,取得了较好的效果.

In all active control methods for the vibration of flexible structures, independent modal space control(IMSC) is an appropriate method to control the vibration of the structure. But it may bring about the observation-spillover and the control-spillover for the structure if an appropriate controller has not been chosen. The neural network control enjoys better robustness, which suppresses the spillover efficiently. In this paper, a neural network controller based on independent modal space control(IMSC) is designed. Simulation results on flexible beams and plates show its effectiveness.