摘要
为了探讨往复式压缩机的故障诊断技术 ,根据专家经验及现场诊断实例 ,将往复式压缩机气缸活塞系统的常见故障分为 10种 ,并对这 10种故障工况下的压缩机指示图进行了归一化处理 ,得到了压缩机指示图典型工况模式图谱。借助功能强大的 MATL AB语言系统及其工具箱 ,完成了 BP网络压缩机工况模式识别软件的设计 ,用实测指示图验证了所建立的神经网络识别模型的准确性。
The common faults in the cylinder and piston system of a compressor are divided into ten kinds based on the expert experience and the diagnosis examples. A PV-diagram atlas of the compressor under typical faulty working conditions are obtained by the unification of the PV-diagrams of the ten faulty conditions. A design of the BP neural networks software for the identification of working conditions of the compressor is completed with the aid of the MATLAB. The accuracy of the networks is proved by measured PV-diagrams.
出处
《振动．测试与诊断》
EI
CSCD
2003年第3期217-219,共3页
Journal of Vibration,Measurement & Diagnosis
