分形理论在装甲车辆滚动轴承故障诊断中的应用

Application of Fractal Theory in Fault Diagnosis of Armored Vehicle Rolling Bearing

滚动轴承由于其工作过程受到载荷、安装、润滑状态等因素的影响,会产生与构件特征频率有关的外圈、内圈及滚动体故障。由于滚动轴承的时域波形在一定的时域长度下存在自相似性,可用分形理论来描述其时间序列的不规则度。考虑到滚动轴承故障信号的特殊性,进行了短时分形维数与分形滤波研究,探讨了滤波器的滤波效果与模糊控制参数间的关系,获得了滚动轴承在常见故障状态下的网格维数和网格维数距离。提出了利用分形维数自动调整模糊控制参数的自适应动态调整方法,有效地改善了噪声滤除效果。针对装甲车辆滚动轴承的正常状态、外圈裂纹、内圈裂纹、滚动体损坏和待测信号5种状态进行了分析与诊断,得到了较好的诊断结果。

Rolling bearing will produce outer race, inner race and rolling body faults related to component character frequency because its work process suffer from the influence from the load, fixing, lubrication state and so on. Because time-domain waveform of the rolling bearing exists self-comparability under definite time-domain length, the abnormity of time sequence can be described by use of fractal theory. To take account of particularity of rolling bearing fault signal, the short-time fractal dimension and fractal filter technology were studied, the relationship between the filter processing results and its fuzzy control parameters was discussed, and gridding dimension and gridding dimension distance under the common fault state of rolling bearing were obtained. A kind of new method was presented. This method is a kind of self-adaptive dynamic adjustor method by use of fractal dimension to automatically adjust the value of the filter fuzzy control parameter, it effectively improves the noise filter effect. Aimed at five kinds of state of armored vehicle rolling bearing such as normal state, outer race crack, inner race crack, rolling body damage and test signal, the vibration signals under these five states were analyzed and diagnosed. The results showed that this method is preferable.