【摘要】 首先阐述了ＣＭＡＣ神经网络的原理、结构和学习算法，提出了一种新的采用竞争学习原理的非等距自适应量化算法。理论分析和仿真结果表明，在保持量化总级数不变的前提下，该方法能有效的利用ＣＭＡＣ神经网络的内存，提高逼近精度。文中进行了理论分析，仿真实验证明了所提方法的有效性。更多还原

【Abstract】 We first discuss the structure and principle of the CMAC neural network. Using competitive learning, we develop a new adaptive quantization algorithm. Theoretical analysis and simulation results show that the CMAC neural network with the new quantication algorithm can use the memory more efficiently and approximate the plant more accurately than the original model.更多还原