The Diagnostic Identification Of Engine Cylinder Wall Clearance Based On BP Neural Network

As engine is the most core part of automobile, it is very important to ensure its bestoperation condition. However, if the technical conditions of the engine can be judged whenit is even not disassembled, it will be of great significance to the technical inspection of theengine.This paper firstly analyzes the inner vibration principles of cylinder of DA462typegasoline engine and its influence on the vibration signals on the surface of the cylinderblock, then makes improvement to test bench of DA462type engine with the principles,and thus more accurate collection to the vibration signals of the engine. Selecting theDenmark B&K company’s PULSE vibration test analysis instrument, collecting differentcylinder wall clearance under the engine block surface vibration signal at the same speed.At the same time, we addition of the Engine drag down experiments in order to exclude allthe effects combustion excitation has on the collected vibration signal, the motor is drivenby the engine and they operate under the same rotational speed, and in the same method onthe same site of the vibration signal acquisition. And this vibration signal reflects the indexof the impact energy piston does to the cylinder wall during compression stroke for thepoint change. and then perform the contrast analysis and calculate the collected vibrationsignals in the time domain, frequency domain and power spectral. At last, after the intervalwavelet packet is used to extract the eigenvectors of vibration signals, part of eigenvectorsare input into different BP neural networks for training and a classifier is established, theother eigenvectors are input into the established classifier for test, and a BP neural networkmodel which is fit for identifying engine cylinder wall gap is put forward. And thefollowing conclusions:With this experimental program, which can effectively exclude the influence ofburning incentive on the vibration signals, the main incentive causing obvious change ofthe vibration signals when the cylinder wall gap enlarges is proved to be the crosswiseimpulse of pistons, and it is also verified that the vibration signal of the engine is a kind ofunstable signal with wide band.The combination of interval wavelet technology and Levenberg-Marquardtoptimization algorithm of BP neural network for diagnosis and recognition of engine cylinder wall gap has achieved ideal results.