Separation Method of Impulsive Fault Component for Gasoline Engine Based on Acoustic Signal Analysis

Abstract

This paper presents a developed dislocation superimposed method (DSM) for automatically extracting the component of impulsive signals from abnormal noise signals of an engine at a single speed range on the basis of the initial DSM. This method consists of three steps: using a correlation analysis to select an appropriate starting superposition point, superimposing abnormal sound signals to improve the signal-to-noise ratio, and intercepting superimposed signals to separate the fault component. Experimental results show that the developed DSM can effectively extract the fault characteristics of cylinder knocking and connecting rod bearing knocking. The developed approach can be applied to separate the fault characteristics of other types of rotating machines.

Adaptive Noise Reduction Method of Synchronous Hydraulic Motor Acoustic Signal Based on Improved Dislocation Superposition Method

“…When the signal is quasiperiodic, the starting point and the superimposed length of each SC needs to be manually selected. Ning et al [3] used the improved DSM to extract automatically the fault components of an engine's quasi-periodic acoustic signal. However, this method utilizes the encoder to position accurately the relative position of the acoustic signal with the rotation period, and the synchronous hydraulic motor cannot install the encoder.…”

“…When the position of the starting point of the actual signal changes, the relative position of the effective information of each SC in the signal period changes. The DSM processing effect in the quasi-periodic signal is related to the actual signal start position [3], and the optimal starting position of the actual signal needs to be determined. In theory, the larger the number of superpositions K in the DSM algorithm, the more accurate the waveform is [8], but this condition will increase computation time.…”

“…The mechanical equipment acoustic signal has three categories of components, namely, the acoustic signal of the component to be analyzed, other components, and background noise. The acoustic signal of the component to be analyzed is usually referred to as a valid signal, and the other two are collectively referred to as noise [3].…”

“…However, under the condition of non-stationary and quasi-periodic signals, determining the starting point of each signal cycle manually is necessary. Ning et al [3] used the improved DSM to extract automatically the fault components of engine quasiperiodic acoustic signals with excellent results. However, this method needs to use the pulse data of the encoder to find the starting position of each signal cycle accurately.…”

Research on Feature Extraction Method of Engine Misfire Fault Based on Signal Sparse Decomposition

“…Because of its simple structure, output power stability, and convenient installation, the engine is widely utilized as the power output device in many mechanical systems, such as automotive, steamship, and spacecraft [1]. However, after the long-term harsh operation condition, the engine and other transmission system are prone to fault which will cause a sudden power interruption, thus affecting the reliability of the whole transmission system [2][3][4]. Furthermore, the vibration signals acquired from the sensor contain not only engine operating information but also various other strong noise signals, which causes useful feature signals to be submerged and makes it hard to diagnose engine fault.…”