The Key Technology Study On Fault Diagnosis System Of I.C. Engine Based On Angular Vibration Of Crankshaft

With extensive applications of internal combustion engine (ICE) and developments of modern science and technology, the conventional breakdown maintenance and periodic monitor methods are not adapted to the demands of present maintenance system. Because of the complexity of ICE, the general mechanical fault diagnosis techniques are difficult to be applied in the diagnosis of ICE, so the condition monitoring and fault diagnosis techniques must be developed to fit in the characteristics of ICE. The crankshaft angular vibration signal of ICE gains abroad attentions both here and abroad for its advantages of simple measurement and abundant information, which is the hotspot in fault diagnosis field of ICE presently. This paper presents the A/D sampling fitting method, the distorted point filtering method and the angular vibration simple harmonic diagnosis method respectively directing towards the key problems existing on signal measurement, signal processing, diagnosis methods, which establish the theoretical basis for studying practical and effective crankshaft angular vibration diagnosis system of ICE.Firstly, following the virtual instrument design principle of software added universal hardware instead of special hardware, it puts forward a new measuring method of angular vibration based on universal A/D sampling card—A/D sampling fitting method which reduces the hardware cost, increases commonality, extents sampling time and can realize the synchronous accurate measurement of angular vibration signal with other signals. A/D sampling fitting method, through adjusting the sampling frequency and data length, can sample smartly the operating cycle number of ICE, angular vibration of unstable operating condition of ICE was measured and analyzed. It also reveals an essential property that there is trend term in angular vibration signal caused by unstable operating condition of ICE, and the effects of trend term on angular vibration analysis and its elimination approaches is studied. When the random measuring error of angular vibration resulted by different reasons is relatively large, it can be eliminated by using the A/D sampling fitting refinement algorithm—multi-point average in one cycle method.Through analysis on the basic characteristic of inference signal in angular vibration signal of crankshaft, a distorted point filtering method (DPF) is presented that only attenuates the errors of distorted point but not affects the real value of non-distorted point in periodic signal. Also it has demonstrated the critical steps of DPF, made a deduction the attenuation rule of the error value of distorted point, and estimated the calculated amount of DPF. Furthermore in order to adapt the calculation speed demands of online measurement or real time control, it also bringsforward fast DPF method. The DPF method and the fast algorithm can be used not only in signal filtering of crankshaft angular vibration signals, but also in other periodic signals and damped harmonic signals. Moreover, for the sake of filtering out the distorted point in damped harmonic signal, it puts forward an idea that changing the damped harmonic signal to simple harmonic signal firstly and then filtering out the distorted point in damped harmonic signal using DPF method. The negative damping Fourier transformation method presented in the paper can be used to transform the damped harmonic signal to simple harmonic signal. It is also a new precise mode parameter identification method.Based upon the dynamics of ICE and Fourier transformation theory, it has studied in depth the relationship between the input excitation moment and the output angular vibration response of crankshaft, and firstly deduced the function relationship of simple harmonic angular vibration of arbitrary measurement point with respective simple harmonic moment of each cylinder in multi-cylinder ICE, namely the basic relationship. Aiming at the present limitations of the rigid and elastic model, by virtue of the orthogonality of simple harmonic vibration, the single order quasi-rigid body model is presented from the point of view of simple harmonic analysis firstly, and two cases satisfying the condition of single order quasi-rigid body model is discussed. Single order quasi-rigid body model is the accurate and simple and practical angular vibration diagnosis kinetic model. Accord to the basic relationship and single order quasi-rigid body model, five deductions are deduced which can be applied in fault diagnosis, and a new angular vibration simple harmonic diagnosis method is established which has the advantages of powerful commonality, accurate diagnosis, simple application, independent of configuration parameters of crankshaft or a great deal of experiments. The new diagnosis method is applied successfully in operation condition monitoring, the misfire cylinder diagnosis, the misfire cylinder rate diagnosis, the gas leakage cylinder diagnosis and the operation uniformity diagnosis of each cylinder.All new methods in this paper are not only deduced theoretically and proved strictly, but tested and verified sufficiently by a great deal of simulated and experimental studies.The research efforts develop the key technology of fault diagnosis system of ICE based on crankshaft angular vibration, has use values on the diagnosis system. Furthermore some generalized meaning research results can also be used to resolve the problems in other relevant fields, and enriches the theories of ICE dynamics, engineering signal processing, measurement techniques, parameter identification of vibration mode and other interrelated subjects.