机械异常磨损微粒在线监测机理与微弱混迭信号辨识方法研究

项目名称： 机械异常磨损微粒在线监测机理与微弱混迭信号辨识方法研究

项目编号： No.51475044

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 机械、仪表工业

项目作者： 郑长松

作者单位： 北京理工大学

项目金额： 83万元

中文摘要： 油液中铁磁性颗粒在线监测是大型机械设备故障诊断的一项共性关键核心技术，电感式在线监测传感器是目前国内亟待研制的监测器。为突破微小颗粒监测机理与多颗粒通过时产生的混迭信号辨识关键技术，本项目拟对其基础理论问题开展研究：1）建立电感式传感器仿真分析数学模型，引入激励线圈频率特性，揭示多参数对感应电动势大小的影响规律；2）研究微小铁磁性颗粒通过传感器的感应电动势特征，自适应匹配激励线圈的激励特性，获得高的感应电动势输出的激励特性与多参数自适应匹配方法；3）研究非规则状态下磨损颗粒通过传感器时产生的微弱信号特征，利用仿真、实验手段分析外界干扰产生的干扰信号特征，采用锁相放大技术进行信号提取，获得微弱信号的提取方法；4）研究处于混沌状态的颗粒通过传感器时所产生的微弱信号特征，利用仿真、实验分析微弱信号混迭特征，建立信号数据库，采用神经网络进行颗粒识别，为油液中微小颗粒在线监测系统提供关键技术支撑。

中文关键词： 信号处理；动态监测；异常磨损

英文摘要： On-line monitoring ferromagnetic particles in the oil is a common key core technology on large mechanical equipment fault diagnosis, inductive online monitoring sensor is urgently needed to be developed at home. To break through key technologies of the tiny particles monitoring mechanism and the identification of mixed overlapping signals produced by many particles passing through the sensor, the basic theory problems are studied: 1) Inductive sensor simulation analysis mathematical model is set up, incentive coil frequency characteristic is introduced, the effects of parameters on the induced electromotive force are revealed; 2) Induced electromotive force characteristics of tiny ferromagnetic particles passing through sensor, incentive characteristics adaptively matching incentive coil, incentive characteristics with multiple parameters adaptively matching methods aiming to obtain high induced electromotive force output are studied; 3) The weak signal feature when wear particles pass through the sensor in an unnormal state is studied, the jamming signal interference characteristics are analyzed by means of simulation and experiment, signals are extracted by lock-in amplifier technology and the weak signal extraction method is obtained; 4) The weak signal feature produced by particles passing through the sensor in a chaotic state is studied, the mixed weak signal characteristics are analyzed by means of simulation and experiment, signal database is established, the neural network is used to identify the particles, the research provides key technical support for the system of on-line monitoring tiny particles in the oil.

英文关键词： signal analysis;dynamic monitoring;exceptional wear