Research On The Rolling Bearings Fault State Recognition Technology Of Based On SOM

Rolling bearing is a kind of important rotating machinery,it has been used in many different fields today. Because it always works in difficult environment,rolling bearing is one of the easiest broken parts in equipment.The damage of rolling bearing often leads to the whole equipment broken,sometimes it also leads to more serious consequences.because of the lifetime discreteness,so it's important to identify rolling bearing status and fault type.An improve EEMD method is used to decompose rolling bearing vibration signal.Then get some Intrinsic Mode Functions.Calculating main IMF energy distribution as rolling bearing fault feature. In this method.Parameter selecting is based on the vibration signal and the added white noise,so it can avoid personal error in traditional EEMD.Because Traditional SOM has long training time and fault type identify accuracy isn't high enough.So immune genetic algorithm is used to optimize the weight updating process.IGA optimize the process by imitate biological immune mechanism and genetic rules. The weight updating process is more comprehensive.So SOM has a stronger ability in data classification.Then put the rolling bearing fault characteristic into SOM which is optimized by immune genetic algorithm.Identify the rolling bearing fault type by IGA-SOM. Finally,using the practical data show IGA-SOM can identify rolling bearing fault type faster and more accurately.