基于IMFECF的多信息熵融合技术的齿轮故障诊断方法

Research of Gear Fault Diagnosis Based on IMFECF of Multi-Information Entropy Fusion Method

交变应力和高负载的工作条件会导致齿轮箱故障频发,为了通过振动信号诊断齿轮箱各工况下运行情况,提出了一种基于IMF熵值分类因子(IMF Entropy Classification Factor,IMFECF)的多信息熵融合技术的齿轮故障诊断方法.通过IMFECF量化信息熵中IMF(Intrinsic Mode Function)的表征能力,分离表征能力优异的IMF,采用信息熵提取IMF中的工况特征,由于不同信息熵各有优势,因此利用多信息熵融合体系结构,获得了最优的自适应模糊推理系统.研究结果表明,经诊断模型训练后的诊断误差满足要求,能准确诊断齿轮箱状态.

Alternating stress and high load working conditions will lead to frequent gearbox faults.To diagnose the gearbox operating conditions through vibration signals,a gearbox fault diagnosis method based on IMFECF multi-information entropy fusion technology was proposed.IMFECF was used to quantify the representation ability of IMF in information entropy,and the IMF with excellent representation ability was separated.The information entropy was used to extract the working condition features in IMF.As different information entropies have their own advantages,the optimal adaptive fuzzy inference system was obtained by using the multi-information entropy fusion architecture.The results show that the diagnosis error of the trained model meets the requirements and the gearbox state can be accurately diagnosed.