轴承
軸承
축승
BEARING
2014年
4期
45-49
,共5页
滚动轴承%故障预测%数学形态学%分形维数%极限学习机
滾動軸承%故障預測%數學形態學%分形維數%極限學習機
곤동축승%고장예측%수학형태학%분형유수%겁한학습궤
rolling bearing%fault prediction%mathematical morphology%fractal dimension%ELM
将形态分形维数作为轴承故障预测特征量，形成轴承故障预测特征量序列。同时，为优化极限学习机（ELM）预测模型，综合考虑模型的精度、预测趋势及稳定性，提出一种序列关联度系数及其计算方法，对ELM预测模型进行优化，并利用提取的故障预测特征量序列对模型进行训练。用轴承全寿命数据进行验证，结果表明，形态分形维数的变化情况较好反映了轴承性能退化的过程，改进的ELM预测模型实现了对轴承故障的有效预测，且其精度及稳定性较原始ELM预测模型有一定提高。
將形態分形維數作為軸承故障預測特徵量，形成軸承故障預測特徵量序列。同時，為優化極限學習機（ELM）預測模型，綜閤攷慮模型的精度、預測趨勢及穩定性，提齣一種序列關聯度繫數及其計算方法，對ELM預測模型進行優化，併利用提取的故障預測特徵量序列對模型進行訓練。用軸承全壽命數據進行驗證，結果錶明，形態分形維數的變化情況較好反映瞭軸承性能退化的過程，改進的ELM預測模型實現瞭對軸承故障的有效預測，且其精度及穩定性較原始ELM預測模型有一定提高。
장형태분형유수작위축승고장예측특정량，형성축승고장예측특정량서렬。동시，위우화겁한학습궤（ELM）예측모형，종합고필모형적정도、예측추세급은정성，제출일충서렬관련도계수급기계산방법，대ELM예측모형진행우화，병이용제취적고장예측특정량서렬대모형진행훈련。용축승전수명수거진행험증，결과표명，형태분형유수적변화정황교호반영료축승성능퇴화적과정，개진적ELM예측모형실현료대축승고장적유효예측，차기정도급은정성교원시ELM예측모형유일정제고。
The morphological fractal dimension is considered as the characteristic quantity of the fault prediction for the bearings,and an array for characteristic quantity of the fault prediction for the bearings is formed.In order to improve ELM prediction model,the precision,trend predication and stability of model are comprehensively considered.An ar-ray relevance factor and its calculation method are proposed.The ELM predication model is improved,and an array for the characteristic quantity of the fault prediction is extracted to train the model.The method is validated through the whole life data for bearings.The results show that the changes in morphological fractal dimension well reflects the process of performance degradation for bearings ,the effective prediction for fault of bearings is realized by improved ELM prediction model,and the precision and stability are improved than those of original ELM prediction model.
