On-line Fault Diagnosis And Analysis Of Rolling Mill Hydraulic Cylinder Based On Acoustic Emission

In the field of iron and steel metallurgy,the on-line fault diagnosis of mechanical equipment is great significance,because this kind of diagnosis can not only greatly save the cost of maintenance,maintenance and replacement of equipment,but also can reduce the probability of production accident caused by equipment failure.However,the research shows that the research of mechanical equipment fault diagnosis for iron and steel industry at home and abroad is still in the initial stage,and the mature theory and method have not been found to be popularized and applied.The hydraulic cylinder of rolling mill is one of the important mechanical equipment in the process of iron and steel production.The failure rate is high.The method of regular maintenance and replacement is often adopted to ensure the stable operation of the hydraulic cylinder of the rolling mill.The research on the on-line fault diagnosis of the hydraulic cylinder of rolling mill is of great significance to the early fault prediction,the reduction of the maintenance cost and the avoidance of production accidents,because the rolling mill hydraulic cylinder is light in reducing the quality of the product,and the weight is caused by the shutdown accident.However,because the fault diagnosis and identification of hydraulic cylinders involve multiple state variables with larger similarity,it is impossible to directly use the pattern recognition method facing the concept of fault management.Therefore,it is of great theoretical significance and practical application value to study the on-line fault diagnosis of rolling mill hydraulic cylinders.In this paper,the theory and method of on-line fault diagnosis for rolling mill hydraulic cylinder based on acoustic emission technology are proposed.Acoustic emission signals are detected on line through acoustic emission detector,and the noiseelimination method and feature recognition method of the state acoustic emission signal of rolling mill hydraulic cylinder are deeply studied.The wavelet de-noising method and the feature classification and recognition method of the three dimensional state heat map and the principal component analysis are used,and the feature recognition results are compared with the wavelet feature recognition method.The advantages and disadvantages of the above three signal analysis methods are given.The new methods and ideas are provided for solving the problem of on-line fault diagnosis of hydraulic cylinders.The main research contents are as follows:(1)the leakage of the hydraulic cylinder of rolling mill and the mechanism of acoustic emission signal are analyzed in detail,and the installation position of the sensor is determined according to the movement principle of the hydraulic cylinder of the rolling mill,and the acquisition of the acoustic emission signal of the hydraulic cylinder in two conditions of normal and leakage is completed.(2)aiming at the problem that the conventional filtering method is difficult to deal with the noise elimination of the state acoustic emission signal of the rolling mill hydraulic cylinder,a wavelet threshold de-noising method is proposed,and the time,frequency and time frequency characteristics of the denoised signal are calculated and feature extraction.(3)aiming at the characteristic classification and recognition of the state acoustic emission signal of the rolling mill hydraulic cylinder,the three-dimensional state thermograph method and the principal component analysis method are put forward respectively,and the comparison and analysis are carried out with the wavelet recognition method,and the simulation results are given.The analysis shows that the three dimensional state thermograph method is suitable for the large similarity between states.The principal component analysis is suitable for the analysis of high dimensional feature data of large samples.The wavelet analysis method is simple and efficient,but it is more suitable for the analysis of small sample data.