将神经网络技术应用于船舶控制系统故障诊断专家系统，给出了系统的组成和知识的表达方法；提出了自增长动态网络学习算法，其优点在于学习迅速，能动态建造多层网；给出了神经网络专家系统的推理方法 The neural network technology is applied to fault diagnosis expert system of ship control system, and the composition and knowledge representation of the system are given. A self-growth dynamic network learning algorithm is proposed, which has the advantages of rapid learning and dynamic building of multi-layer network. Inference method of neural network expert system 