简要介绍了火电机组热力系统故障智能诊断系统的总体功能结构和诊断模型构成。结合某30 0 MW火电机组高压给水加热器系统故障诊断实例 ,详细阐述各模块的功能和实现方法。针对模型软件开发中的关键技术 ,详细论述了电站热力系统故障诊断中语义型和趋势型两种征兆的模糊数学表达及合成计算方法 ,该综合运用有利于提高故障诊断的及时性和诊断结果的可靠性 ;给出了模糊模式识别和人工神经网络两种故障智能诊断模型 ,讨论了两种模型的技术要点。给出了高压给水加热器系统具体诊断算例 ,验证了所述技术和方法的有效性 This paper briefly introduces the overall functional structure and diagnostic model of fault diagnosis system of thermal power system. Combining with a 300 MW thermal power unit high pressure feedwater heater system fault diagnosis examples, elaborated the function of each module and its implementation. Aiming at the key technologies in the development of model software, the fuzzy mathematics expression and synthetic calculation method of semantic and trend type symptoms in power plant thermal system fault diagnosis are discussed in detail. The comprehensive utilization is helpful to improve the timeliness of diagnosis and diagnosis Reliability. Two kinds of intelligent fault diagnosis models based on fuzzy pattern recognition and artificial neural network are given, and the technical points of the two models are discussed. A specific example of diagnosis of high pressure feedwater heater system is given, which verifies the effectiveness of the technique and method