单电磁悬浮系统的神经网络模糊滑模控制

为实现单电磁悬浮系统悬浮气隙的精确控制,提出一种基于神经网络的模糊滑模控制方案.根据单电磁悬浮系统的动态非线性数学模型,设计使系统状态在有限时间内到达稳定点的滑模面,同时根据滑模切换状态,通过引入神经网络的模糊控制方法对滑模切换控制量的增益进行评估,实时对滑模控制量进行调整,实现切换控制信号的柔化.基于神经网络的模糊滑模控制系统不仅能很好地跟踪给定信号,而且能削弱滑模控制抖振,对外部扰动具有完全的鲁棒性.仿真结果表明,所设计的控制系统零超调,具有速度跟踪性能,对外部扰动具有很强的鲁棒性.

Neural network fuzzy sliding mode control for single electromagnetic levitation system

In order to actualize the precise control of levitation airgap in the single eletromagnetic levitation system, a fuzzy sliding mode control scheme based on neural network was proposed. According to the dynamic nonlinear mathematical model for the single eletromagnetic levitation system, the sliding surface 〖JP2〗which could make the system state reach the stable point within the finite time was designed. Simultaneously, 〖JP〗according to the switching state of sliding mode, the gain of control amount in sliding mode was evaluated through introducing the fuzzy control method of neural network. In addition, the control amount of sliding mode was adjusted in real time, and the softening of switching control signal was actualized. The fuzzy sliding mode control system based on the neural network can not only track the given signal, but also weaken the chattering of sliding mode control and posses global robustness to the external disturbance. The simulated results show that the designed control system has zero overshoot, fast tracking performance and strong robustness to external variation.