神经网络具有多级、并行、分布式、高度容错能力，并具有自行组织自行发展的适应性功能，能在被处理的信息内容引导下，自行改造其自身的结构及其运行规则，是研究非线性的、适应的、大脑风格的信息处理的全新工具，因此，本文提出一种模糊网络模型和自适应学习算法，用于自然语言的识别处理控制． Neural networks are multi-level, parallel, distributed, highly fault-tolerant, and have the adaptive function of self-organizing self-development. It can self-transform its own structure and operation rules under the guidance of the processed information content. Linear, adaptive and brain-style information processing. Therefore, this paper proposes a fuzzy network model and adaptive learning algorithm for the control of natural language recognition processing.