基于 MEEMD 算法的二冲程柴油发动机机体振动分析
2024-05-21贺献忠徐麟绍高超
贺献忠 徐麟绍 高超



摘要:二沖程低速柴油机具有复杂的振动特性,传统的经验模态分解(EMD)算法对其振动信号处理效果不理想。为此,采用修正集合经验模态分解(MEEMD)算法对低速柴油机机体振动信号进行分解。首先,采用三轴加速度计测量发动机机体振动。然后利用均方根(RMS)对三轴振动强度进行分析。最后,对x轴上的信号进行MEEMD分析。结果表明:砌块在x轴方向的振动强度最大;与EMD算法相比,MEEMD算法可以抑制模态混合,有助于更好地识别块振动激励。
关键词: 低速柴油机 振动 信号处理 修正集合经验模态分解
中图分类号:TM935
Analysis of the Block Vibration of Two-Stroke Diesel Engines Based on
the MEEMD Algorithm
HE Xianzhong1* XU Linshao2 GAO Chao3
1. Merchant Marine College, Shanghai Maritime University, Shanghai, 201306 China;2. Zhejiang Shipping Group Zhehai Shipping Co., Ltd., Hangzhou, Zhejiang Province, 311500 China;3. Transportation Department, Shougang Mining Corporation, Tangshan, Hebei Province, 064404 China
Abstract:The two-stroke low-speed diesel engine has complex vibration characteristics, and the conventional empirical mode decomposition (EMD) algorithm has the unsatisfactory processing effect of its vibration signals, so the modified ensemble empirical mode decomposition (MEEMD) algorithm is adopted to decompose the signals of the block vibration of the low-speed diesel engine. First, a triaxial accelerometer is used to measure the block vibration of the engine. Then, the root mean square (RMS) is used to analyze the vibration intensity of three axes. Finally, the MEEMD algorithm is used to analyze the signals on the x-axis. The results show that the vibration intensity of the block in the x-axis direction is the largest, and that the MEEMD algorithm can suppress modal mixing compared to the EMD algorithm, which helps to better identify the vibration excitation of the block. Key Words: Low-speed diesel engine; Vibration; Signal processing; Modified ensemble empirical mode decomposition
柴油机零部件众多,振动激励源复杂[1]。特别是二冲程低速柴油机具有独特的十字头总成,其振动比四冲程柴油机更复杂。因此,有必要对低速柴油机机体的振动信号进行研究分析。
经典算法包括傅里叶变换和小波变换等[2]。经验模态分解(EMD)算法能够对信号进行自适应分解,避免了傅里叶变换和小波变换依赖窗函数进行分析的缺点[3]。然而,EMD分解结果中存在模态混叠和伪分量[4,5]。为了减少EMD的缺点,改进的集合经验模态分解(MEEMD)算法被提出[6],并已成功应用于各个领域。如对行星齿轮箱的振动信号进行分解[7],可滤除包含主要故障信息的固有模态函数(IMF)分量[8]。MEEMD 算法也被用于地震分析[9-10],研究地震波的变化特征。
上述研究结果表明,MEEMD 算法通过重构本征模函数,具有有效特征识别的可靠方法。然而,尚未应用于船舶低速柴油机缸体的振动分析。……
