【作者】 张凯锋； 戴先中； 张腾； 陆翔；

【Author】 Zhang Kaifeng Dai Xianzhong Zhang Teng Lu Xiang

【机构】 东南大学自动控制系；

【摘要】 文章针对一般的MIMO非线性系统,利用神经网络逆系统理论实现原系统的线性化和解耦。然后对由神经网络逆系统和原系统串联组成的复合伪线性系统,采用“棒棒+阻尼”的非线性控制方法设计附加控制器,以达到时间最优,从而克服传统线性控制方法在处理类似问题时存在的快速性与平稳性、静态精度与动态稳定性的矛盾。最后对两输入两输出非线性系统的仿真结果验证了该方法的有效性。更多还原

【Abstract】 In this paper, based on NN inverse system theory, normal MIMO nonlinear system is linearized and decoupled by cascading its NN inverse system before it. Then, aiming at this pseudo-linear system, accessory nonlinear controller is designed according to "Bang-Bang & Damping" strategy. This method can achieve time-optimal, as well as solve the conflicts between speed and stability, between static accuracy and dynamic accuracy, which are existed when adopting traditional linear control method. To a given example of 2-input 2-output nonlinear system, the simulation result indicates the effectiveness of this method.更多还原