基于稀疏自动编码器和支持向量机的矿渣立磨主减速机故障预警研究

Research on Fault Early Warning of Main Reducer of Slag Vertical Mill Based on Sparse Automatic Encoder and Support Vector Machine

立磨主减速机是矿渣立磨的核心设备,然而由于设备结构和工作条件的复杂性,难以对其进行准确的故障诊断。提出了一种基于小波包、稀疏自动编码器和支持向量机对立磨主减速机进行故障预警和诊断的方法。利用小波包对立磨主减速机的振动信号构建能量特征向量,通过稀疏自动编码器对特征向量进行数据重构,利用重构数据将支持向量机训练成具有多重感知功能的分类器。实验研究结果表明,基于小波包、稀疏自动编码器和支持向量机的故障诊断方法可以有效地对故障进行分类。按照该方法针对立磨主减速机开发故障预警模型,利用立磨主减速机故障时的振动信号,建立故障诊断模型,实现立磨主减速机的故障预警和故障诊断。

The main reducer of vertical mill is the core equipment of slag vertical mill.However,due to the complexity of equipment structure and working conditions,it is difficult to carry out accurate fault diagnosis.This paper presents a method of fault early warning and diagnosis for the main reducer of the vertical mill based on wavelet packet,sparse automatic encoder and support vector machine.The wavelet packet is used to construct the energy feature vector from the vibration signal of the main reducer of the vertical mill.The data of the feature vector is reconstructed by the sparse automatic encoder,and the reconstructed data is used to train the support vector machine into a classifier with multiple sensing functions.The experimental results show that the fault diagnosis method based on wavelet packet,sparse automatic encoder and support vector machine can effectively classify the faults.According to this method,the fault early warning model is developed for the main reducer of vertical mill,and the fault diagnosis model is established by using the vibration signal of the main reducer of vertical mill,so as to realize the fault early warning and fault diagnosis of the main reducer of vertical mill.