发动机表面振动限值的智能确定方法研究  

Research on the Determination of Engine Surface Vibration Limits Based on Intellective Learning Method

作　　者：莫易敏[1] 赵雷刚[1] 潘丽娜[1] 汤春球[1] 

机构地区：[1]武汉理工大学机电学院,武汉430070

出　　处：《武汉理工大学学报》2010年第23期116-119,共4页Journal of Wuhan University of Technology

摘　　要：通过理论和试验分析影响发动机表面振动的发动机运行参数,得出发动机表面振动加速度信号RMS与发动机的转速、扭矩有关。为了实现发动机变工况的基于振动的运行状态监测,提出智能学习的发动机表面振动限值确定方法,该方法在利用振动信号对发动机监测之前,学习一定时间发动机的不同工况振动信号特征值,并剔除同工况下异常的数据,然后利用平均值确定该工况下的振动特征上下限值。通过实例验证表明该方法能够实现对发动机表面振动限值的确定。Through theoretical and experimental analysis,RMS of the engine surface vibration signals is related to engine speed and torque.In order to realize the engine variable condition monitoring based on vibration,a statistical learning method is proposed to determine the engine surface vibration limits.The method learns the engine vibration signal characteristics value in different conditions for some time before engine monitoring and removes the wrong data.Then it determines the upper and lower limits of the vibration characteristics by using the mean of same operating conditions.The result also demonstrates the method is effective on the determination of vibration limits.

关 键 词：统计学习 振动限值 发动机 变工况 

分 类 号：TH113.1[机械工程—机械设计及理论]