【摘要】 精对苯二甲酸(pure terephthalic acid,PTA)是生产聚酯的重要原料。针对大型PTA装置故障的特点,提出将小波分析、模糊ART神经网络和专家系统等技术结合起来,开发了实时故障诊断专家系统。文中阐述了专家系统的具体实现,包括数据预处理模块、知识库组成及表达方式、人机界面、推理机的组成和诊断过程等。在某大型PTA装置中的成功应用表明:该系统具有预报准确、界面友好、易于维护等特点,有很好的通用性和推广价值。更多还原

【Abstract】 Pure Terephthalic Acid (PTA) is an important material for producing polyester.Aiming at the fault characteristics of the large PTA plant,a real-time fault diagnosis system based on wavelet transform,fuzzy ART neural network and expert system is proposed.Pre-processing,the content of knowledge base,user interface,the hybrid reasoning and diagnosing strategy are introduced.Finally the proposed approach is successfully applied to large scale PTA plant for fault prediction and diagnosis.The application results show that it can make accurate forecast and has many specialties such as friendly interface,easy maintenance,high universality and great practical value.更多还原