基于自适应优化品质因子的共振稀疏分解方法及其在行星齿轮箱复合故障诊断中的应用

发布时间：2018-04-03 07:10

本文选题：共振稀疏分解　切入点：品质因子　出处：《机械工程学报》2016年15期

【摘要】：在共振稀疏分解方法中,品质因子决定其共振属性,其值的选择对共振稀疏分解结果有着很大的影响。现有的共振稀疏分解方法主要是依靠人为选择品质因子,带有较大的主观随意性,对最终诊断结果的提升非常有限。为此,基于遗传算法的全局优化性能,提出一种自适应优化品质因子的共振稀疏分解新方法。与已有方法相比,该方法利用遗传算法优良的寻优性能,优化共振稀疏分解中的品质因子,自适应地得到与输入信号故障特征相匹配的高低共振分量的品质因子。将所提出的新方法应用于某行星增速齿轮箱中行星齿轮与行星架轴承的复合故障诊断中,有效地提取出振动信号中相应的故障特征,实现了早期复合故障的准确诊断,表明了该方法的有效性和实用性。
[Abstract]:In the resonance sparse decomposition method, the quality factor determines the resonance attribute, and the selection of its value has a great influence on the resonance sparse decomposition results.The existing resonance sparse decomposition methods mainly rely on the artificial selection of quality factors, with a large subjective randomness, and the improvement of the final diagnosis results is very limited.Therefore, based on the global optimization performance of genetic algorithm, a new resonant sparse decomposition method for adaptive optimization of quality factors is proposed.Compared with the existing methods, this method optimizes the quality factors in the resonance sparse decomposition by using genetic algorithm, and adaptively obtains the quality factors of the high and low resonance components matching the fault characteristics of the input signals.The proposed new method is applied to the composite fault diagnosis of planetary gear and planetary bearing in a planetary gearbox. The corresponding fault characteristics of vibration signal are extracted effectively, and the accurate diagnosis of early composite fault is realized.The effectiveness and practicability of the method are demonstrated.
【作者单位】：哈尔滨工业大学机电工程学院;
【基金】：国家自然科学基金(51175102) 中央高校基本科研业务费专项资金(HIT.NSRIF.201638)资助项目
【分类号】：TH132.41

【相似文献】