Research And Implementation Of A Hammer Crusher Fault Diagnosis System Based On Signal Feature Recognition

In modern industrial production,the single-rotor hammer crusher is a kind of largescale rotating machinery widely used.Due to various random influences,various factors make it inevitable that some parts of the crusher will have failures such as impact and wear.Finally,the whole crusher equipment can not operate normally.Rotor as the core of crusher,its state fault detection is of vital importance.In this paper,a crusher fault diagnosis device is designed based on the data from the test bench.The main work of this paper is as follows:1.An improved noise reduction algorithm for ST-SVD was designed and developed.For the acquisition environment of rotor vibration signal,S transformation and singular value decomposition are used to for de-noising.Hankel matrix is obtained by S transformation and the threshold is determined by differential spectrum method.The experimental results show that this method can effectively remove the noise and obtain a good well-reconstructed signal.2.An EMD method based on crusher rotor vibration signal feature extraction method has been designed.Firstly,we did time domain and frequency domain analysis on the vibration acceleration signals of rotor,vibration displacement signal and speed signal in order to find out the optimal signal of fault characteristics of category.Then,after reusing EMD for feature extraction,we found out the optimal IMF component,thus choosing the most obvious and the most sensitive characteristic.3.A system of rotor fault recognition based on SVM and PSO was designed.In terms of SVM parameter selection and high-dimensional disaster,we selected PSO algorithm for optimization,and the experimental data and the previously extracted eigenvalues for verification.Finally,we used data from crusher to reverify them.The results show that the algorithm can effectively reduce the error and improve the accuracy of prediction.?...