GBF-CMAC和滑模控制的柔性结构系统控制

针对一类不确定系统的跟踪控制，设计了一种将GBF-CMAC（cerebellar model articulation controller with Gauss basis function）与滑模控制相结合的控制系统。利用符号距离和分层结构减少了神经网络所需存储器的数量，并提出了一种神经网络参数的自适应学习律。将设计的控制器用于含有不确定性和欠驱动结构的高阶柔性直线结构系统的跟踪控制，并与一般滑模控制和积分滑模控制进行了比较。实验结果表明，所设计的控制器不仅具有较好的鲁棒性，而且改善了滑模控制存在的抖振问题。同时通过调整神经网络的参数对抖振进行控制，实现了抖振和跟踪性能之间的最优选择。

Flexible plant system control based on GBF-CMAC and sliding mode control

In this paper, a tracking control method combining cerebellar model articulation controller with Gaussian basis function (GBF-CMAC) and sliding mode control (SMC) for uncertain systems is designed. An adaptive learning algorithm of GBF-CMAC is proposed, in which a signed distance and hierarchical structure are used to reduce the memory capacity needed by the neural network. The designed controller is applied for the tracking control of high-order flexible linear system with uncertainties and under-actuated structures, and it is compared with general SMC and integral sliding mode control (ISMC). The experimental results show that the designed controller has a better robustness and improves the chattering problem of SMC. Moreover, the chattering is controlled by adjusting the parameters of the neural network to achieve the optimal choice between chattering and tracking performance.