针对风力机振动信号采集过程中易受噪声影响的问题,提出基于过完备原子库的匹配追踪算法对风机振动信号进行处理。该算法能自适应提取和原子相关的信号结构,从而可实现噪声抑制。在匹配追踪算法处理过程中,利用结合梯度信息的改进的粒...针对风力机振动信号采集过程中易受噪声影响的问题,提出基于过完备原子库的匹配追踪算法对风机振动信号进行处理。该算法能自适应提取和原子相关的信号结构,从而可实现噪声抑制。在匹配追踪算法处理过程中,利用结合梯度信息的改进的粒子群优化算法来寻找最佳原子。仿真结果表明,该算法比标准匹配追踪算法具有更快的运算效率及更高的重构精度。利用该算法对风力发电机齿轮箱振动信号进行去噪处理实验。实验结果表明,去噪后信号信噪比可提高5 d B以上,波形特征更加清晰,并且可以在降噪的同时有效保留故障信息。展开更多
互补集合平均经验模态分解(complementary ensemble empirical model decomposition,CEEMD)作为一种时频特征分析方法,可以较好地提取复杂非线性非平稳信号的故障特征,但其存在虚假分量,很大程度限制诊断过程中的准确性。针对该问题,提...互补集合平均经验模态分解(complementary ensemble empirical model decomposition,CEEMD)作为一种时频特征分析方法,可以较好地提取复杂非线性非平稳信号的故障特征,但其存在虚假分量,很大程度限制诊断过程中的准确性。针对该问题,提出一种基于KL散度(Kullback-Leibler divergence,KLD)的CEEMD虚假分量识别方法(KL-CEEMD)。该方法在原有CEEMD方法基础之上,进一步计算各分量IMF与原信号之间的KL散度值,从而量化各分量与原信号之间的相关性。最后通过对各个IMF的KL散度值进行聚类分析,找出虚假分量和真实分量,最终解决CEEMD的虚假分量问题。为验证KL-CEEMD的有效性,研究搭建风力机传动系统振动试验台,基于该方法对实验台实验数据以及仿真数据进行验证性研究,最终证明所提方法可以很好改善CEEMD的虚假分量问题,能够有效提取出故障信号的真实特性。展开更多
The impeller of turbo machinery is a typical nonlinear multi-oscillator system.The vibration of each module is coupling, including fluid-solid coupling of the blade.The subject of our investigation was a KDF-5 mine fa...The impeller of turbo machinery is a typical nonlinear multi-oscillator system.The vibration of each module is coupling, including fluid-solid coupling of the blade.The subject of our investigation was a KDF-5 mine fan for which we analyzed air vibration signals and axial vibration signals by using correlation dimension analysis under five variable working conditions.The results indicate that their correlation dimension curves show a uniform trend.That is to say, the characteristics of the variation signals of the integral structure are correlated and mutually embodied.It proves that it is possible to monitor the working state of a mine fan by coupling the vibration signals and air vibration signals for these are more sensitive in representing the status of the impeller system.展开更多
文摘针对风力机振动信号采集过程中易受噪声影响的问题,提出基于过完备原子库的匹配追踪算法对风机振动信号进行处理。该算法能自适应提取和原子相关的信号结构,从而可实现噪声抑制。在匹配追踪算法处理过程中,利用结合梯度信息的改进的粒子群优化算法来寻找最佳原子。仿真结果表明,该算法比标准匹配追踪算法具有更快的运算效率及更高的重构精度。利用该算法对风力发电机齿轮箱振动信号进行去噪处理实验。实验结果表明,去噪后信号信噪比可提高5 d B以上,波形特征更加清晰,并且可以在降噪的同时有效保留故障信息。
基金Projects BK2005018 supported by the Natural Science Foundation of Jiangsu Province CX07B-061z by the Graduate Research and Innovation Plan of Jiangsu Province
文摘The impeller of turbo machinery is a typical nonlinear multi-oscillator system.The vibration of each module is coupling, including fluid-solid coupling of the blade.The subject of our investigation was a KDF-5 mine fan for which we analyzed air vibration signals and axial vibration signals by using correlation dimension analysis under five variable working conditions.The results indicate that their correlation dimension curves show a uniform trend.That is to say, the characteristics of the variation signals of the integral structure are correlated and mutually embodied.It proves that it is possible to monitor the working state of a mine fan by coupling the vibration signals and air vibration signals for these are more sensitive in representing the status of the impeller system.
