基于LabVIEW与知识工程的故障诊断系统设计

Design of Fault Diagnosis System Based on LabVIEW Combined with Knowledge Engineering

为了利用多源信息、融合多种评判结果,设计利用性能参数、热力学参数进行模糊诊断,基于LabVIEW提取振动信号的小波包子带能量特征进行神经网络模式识别,领域专家通过权值系数对诊断策略进行调整和评估,由模糊评价进行综合诊断的系统框架,在应用中取得了较好的效果。

To use information of different sources and mix multiple results, a scheme of using performance character and thermodynamics parameter is proposed. In this scheme, the wavelet packet sub-band energy of vibration signal is achieved based on LabVIEW and the neural net pattern reorganization is implemented based on performance character. Experts of each domain adjust the diagnosis strategy using weight coefficient, and the final conclusion is combined through fuzzy evaluation. The proposed method has been used successfully.