Research On Detection Method Of Engine Cylinder Wall Clearance Based On Empirical Mode Decompostion

With the rapid development of China’s economy and the improvement of the people’s living standards,the number of cars’ social possessions continues to increase.At the same time,the volume of transactions in the used car market is also increasing.As the engine’s power output source,the engine’s technical condition will directly affect the running status of the vehicle.Therefore,the detection and fault diagnosis of engine technology conditions are even more important.The existing engine fault diagnosis system is mainly designed and developed for the engine electronic control system.There are no mature detection technologies and equipment for the detection of technical conditions of engine mechanical components(such as cylinder wall clearance,bearing clearance,etc.).Measurements and calculations are still required after the engine is disassembled,which is not only time consuming,laborious,but also destroys the normal fit of the engine.Based on the analysis of the working principle and vibration mechanism of the engine,the paper collects and analyzes the surface vibration signals of the engine cylinder block to achieve rapid and accurate detection of the engine cylinder wall gap without disassembling the body to compensate for the current situation.Inadequacies in the detection of the technical condition of engine mechanical components.This article takes the DA-462 gasoline engine as the main research object and sets up the engine vibration signal acquisition and testing system together with the vibration signal acquisition equipment.Artificial simulation of the engine in the cylinder wall clearance of0.03 mm,0.06 mm,0.09 mm,0.12 mm,0.15 mm and 0.18 mm operating conditions,in each cylinder wall clearance state speed was measured respectively 800r/min,1500r/min and2200r/min vibration signal.For the characteristics of the engine vibration signal,and according to the wavelet transform has an excellent effect on signal denoising,the set of empirical mode decomposition has a good self-adaptiveness for non-stationary signals.Combining the advantages of the two methods,a signal processing method based on wavelet-set empirical mode decomposition is proposed to analyze and process the engine vibration signal,and the characteristic signal in the signal is expressed using approximate entropy,and the characteristic signal is identified using a neural network.Finally,the clearance of engine cylinder wall is verified.The analysis verifies that the method based onwavelet transform and ensemble empirical mode decomposition has good effect on the detection of engine cylinder wall clearance and provides a reference for the engine’s detection of mechanical dislocation.