混合系统CMAC直接逆模型控制

Direct Inverse Model Control of Hybrid System based on CMAC Neural Network

CMAC神经网络是一种具有线性结构、算法简单的局部逼近网络。考虑混合系统中参数不确定等因素影响 ,研究了一类混合系统的 CMAC神经网络控制器的设计 ,采用直接逆模型控制方式 ,不仅可以缩短网络训练时间 ,而且实现了一类混合系统的快速跟踪控制。

CMAC neural network is a kind of local network with linear structure.Based on CMAC neural network,direct inverse model control is put forward to control hybrid dynamical system.The proposed control method makes the design of neural controllers more feasible and practical for hybrid dynamical system,since it greatly shortens the quality and quantity requirements of the teaching signals,and reduces the long training time of neural network.Simulation results indicated that the proposed control method is very effective in hybrid systems.