水电机组状态检修实现策略与框架设计

给出了水电机组状态检修系统的逻辑结构;针对水电机组故障多样性和复杂性，重点提出了一种由基于数据融合的模糊神经网络技术和专家系统的在线诊断子系统及基于模型的离线诊断子系统组成的故障诊断系统，利用专家系统的逻辑推理能力以及人工神经网络的学习和容错能力，建立转子非线性动力模型仿真来离线分析那些没有办法定量提取或解释的征兆，对机组进行故障诊断;同时给出了实时数据的一种智能化存储策略，很好地解决了在线监测所产生的海量数据。

IMPLEMENTATION STRATEGIES AND FRAMEWORK DESIGN OF HYDRO-GENERATING UNIT CONDITION MAINTENANCE SYSTEM

Introduced is the logistic construction of the hydro-generating unit condition maintenance system. Especially, in accordance with the diversity and complexity of hydro-generating unit faults, a type of fault diagnosis system based on multi-diagnosis methods is presented, which consists of two sub-diagnosis systems: one is an on-line sub-diagnosis system based on the data-amalgamation neural network (NN) and expert system (ES); the other is a model-based off-line sub-diagnosis system, which makes the best of the logic reasoning faculty of ES, and the learning and fault-containing ability of NN. The symptoms can be diagnosed that cannot be rationally picked out or explained just based on non-linear simulation of rotors and analysis of bearing conditions. An intelligent data store strategy is given to store the real-time data from the on-line monitoring system.