摘要
针对板带轧机液压ＡＧＣ系统，根据专家系统基本原理，结合古典信号处理方法、模糊理论、神经网络理论，建立了系统故障诊断系统的结构形式和学习算法。利用模糊诊断理论进行模糊推理以解决系统故障的实时诊断，利用神经网络对模糊推理模型进行训练以提高诊断的准确率，并可对未知的知识进行学习和补充。开发了液压厚度自动控制（ＡＧＣ）系统故障诊断专家系统软件，通过实验证明所用方法有效。
For the hydraulic automatic gauge control (HAGC)system of the strip rolling mill, the framework of fault detecting and diagnosis (FDD)expert system is set up based on the ultimate of expert system, compound with the classical method of signal processing, fuzzy theory and neural networks theory. Where,using a neural network to train the fuzzy reasoning model,then making use of this model to diagnose the faults of the system, the real time fault diagnosing for the system can be realized, the diagnosing precision can be enhanced, and the new information for the expert system can be renewed. A set of FDD software for the hydraulic automatic gauge control (AGC)is developed. The approach is approved in effect via the experiment.
出处
《液压气动与密封》
2000年第1期29-31,共3页
Hydraulics Pneumatics & Seals
