【摘要】 随着非线性科学理论的飞速发展,信息熵方法已广泛应用于工程机械的故障诊断中。提出了层次分析和离散熵结合的层次离散熵(HDE),研究其参数变化对熵值计算精度和计算效率的影响,并与多尺度熵(MSE)、多尺度模糊熵(MFE)、多尺度离散熵(MDE)、层次样本熵(HSE)和层次模糊熵(HFE)进行对比研究,研究结果表明,HDE在鲁棒性和计算效率方面有一定的优势。基于HDE和成对邻近特征选择方法(PWFP)的优点,提出了基于HDE和PWFP的高压共轨喷油器故障诊断方法,通过对共轨柴油机燃油压力波数据进行分析,并将该方法与MSE, MFE, MDE, HSE, HFE和PWFP结合方法作了对比。结果表明,相对于现有信息熵的故障诊断方法来说,基于层次离散熵和PWFP的喷油器故障诊断方法有更高的故障识别率和计算效率。更多还原

【Abstract】 With the rapid development of nonlinear scientific theory, the information entropy method has been widely used in fault diagnosis of engineering machinery. In the paper, hierarchical dispersion entropy(HDE) based on the combination of analytic hierarchy process and dispersion entropy was proposed and the influence of parameter variation on the calculation accuracy and computational efficiency of the entropy value was studied. The comparison of HDE with MSE, MFE, MDE, HSE and HFE was carried out, and the results show that the robustness and computational efficiency of HDE have certain advantages. By virtue of the advantages of HDE and the feature selection method PWFP, a fault diagnosis method for high pressure common rail injectors based on HDE and PWFP was proposed. The fuel pressure wave data of a common rail diesel engine were analyzed, and the method of PWFP combined with MSE, MFE, MDE, HSE or HFE was also tried. It is shown that compared with the existing information entropy fault diagnosis method, the fault diagnosis method based on hierarchical dispersion entropy and PWFP proposed in the paper has higher fault recognition rate and calculation effectiveness.更多还原