关键词: 三维微重力;地面模拟系统;神经网络;滑模控制, 随机风 高速列车 非定常气动载荷 功率谱密度

Abstract: Difficult problems of simulation the space microgravity environment for big and middle experimental objects with complex move are considered. A novel system is proposed and the adaptive compensation control method based on neural sliding mode is designed. The equipments combine machine transmission、motor drive and air-bearing to realize three-dimension spatial microgravity simulation. Force feedback controller counteracts gravity real timely as controller. The scheme dispenses with accurate dynamical model. Friction non-linearity of machine system and outside disturbance are considered. Controller is designed based on state feedback. Sliding model controller is employed to compensate uncertainty model to guarantee robust. RBF neural network is used to adaptive study uncertainty up bound to reduce “chattering”. The stability is proved based on Lyapunov theory. The simulation shows that the design of test system can achieve higher simulation accuracy, has important engineering value.

Key words: three-dimensional microgravity;ground simulation system;neural network;sliding model control, Stochastic wind High-speed train Unsteady aerodynamic loads Power spectral density