Fault Diagnosis Method of Diesel Engine Injector Based on Hierarchical Weighted Permutation Entropy

Abstract

In order to solve the shortcoming of multi-scale weighted permutation entropy ignoring high-frequency components, this paper proposes a new nonlinear dynamics analysis method, which is called hierarchical weighted permutation entropy. This method can fully and accurately reflect the low-frequency and high-frequency information of the time series. Based on this, a new intelligent fault diagnosis method for common rail injector based on hierarchical weighted permutation entropy and LSSVM is proposed. Through simulation signal and experimental data analysis, the proposed method is compared with the fault diagnosis method based on multi-scale permutation entropy and multi-scale weighted permutation entropy. The results show that HWPE can more comprehensively reflect the fault information and the proposed diagnosis method has a higher fault recognition rate and stronger robustness.