论文部分内容阅读
广义小脑模型（ＣＭＡＣ）在函数泛化能力和函数逼近能力方面优于基本ＣＭＡＣ模型，但算法较为复杂，实时性差．因此，研究广义ＣＭＡＣ模型的快速算法，对于满足实时控制是非常必要的．文中研究了基于高斯基函数的广义ＣＭＡＣ模型的快速算法，定义了包含待学习样本点的一个超立方体子空间，提出了基于该超立方体子空间的快速学习算法．通过算例仿真表明，学习算法收敛速度较快，可以满足实时控制要求．
The generalized cerebellar model (CMAC) is superior to the basic CMAC model in terms of generalization of functions and approximation of functions, but the algorithm is more complex and has poor real-time performance. Therefore, it is very necessary to study the fast algorithm of generalized CMAC model to satisfy real-time control. In this paper, a fast algorithm of generalized CMAC model based on Gaussian basis function is studied. A hypercube subspace containing the sample points is defined. A fast learning algorithm based on the hypercube subspace is proposed. The simulation results show that the convergence speed of learning algorithm is fast and can meet the real-time control requirements.