“倒三角”系统的神经滑模控制

针对高阶非线性、强耦合倒三角平衡系统,设计一种神经滑模控制器（NNSMC）对其进行平衡控制。所设计的控制律由系统进入滑动模态的等效控制量及神经网络学习算法计算得出的系统补偿控制量构成。其中,等效控制部分保证系统滑动模态可达,补偿控制量用于对系统的干扰和不确定性的补偿。该方法既保留了滑模控制所具有的较强的鲁棒性,又使控制系统滑动模态的品质得到保证和改善,同时削弱了系统的抖振。计算机仿真结果表明了该方法的有效性和可行性。

Neural Network Sliding-mode Controller for a Seesaw System

A neural network sliding-mode controller（NNSMC）designing method was proposed to balance a higher-order nonlinear strong coupling seesaw system.The control law consists of equivalent control and compensation control determined by network learning algorithm.The equivalent control ensures the system to enter the sliding-mode.The compensation control is used to compensate the interference and uncertainties of the system.This method not only retains the strong robustness of system,but also guarantees and improves the quality of the sliding-mode.Chattering of the system is eliminated at the same time.Computer simulation results show that the method is effective and feasible.