将 Kohonen 人工神经网络模型引入专家系统。概述了 Kohonen 模型的基本原理和学习算法——自组织映射,研究了用此模型取代柑桔区域规划专家系统知识库及推理机的适用性和优越性。研究结果表明,该模型能有效地进行知识获取、知识表示及自适应推理,从而可以克服目前大多数专家系统在这些方面存在的困难。 Introduce Kohonen artificial neural network model into expert system. The basic principle of Kohonen model and learning algorithm - self-organizing mapping are summarized. The applicability and superiority of using this model to replace citrus regional expert system knowledge base and inference engine are studied. The results show that the model can effectively acquire knowledge, express knowledge and adaptive reasoning, which can overcome the current difficulties in most of these expert systems.