Browsing by Subject "model identification"

Now showing 1 - 2 of 2

Access status: Open Access ,
Automobile gearbox diagnostics on the basis of the reference model
(2012) Komorska, Iwona
A method of diagnostics of the automobile gearbox wearing out is presented in this paper. The procedure can be automatically realised during driving. To be able to perform such diagnostics the reference model of the vibration signal on the basis of the synchronously averaged signal was created for a new gearbox (in a good technical condition). In order to realize the procedure on-board this model must be easily identifiable. A complicated structural model was given up for the simpler model of the abstract vibrations signal. Three ways of modelling were presented: harmonic series, parametric ARMA (Auto Regressive Moving Average) model, and the wavelet model. The current vibration signal is recorded at the determined moment of driving and then compared with the reference model. After each overhaul or exchange of drive system elements the reference model should be identified again. Diagnostics is being done by comparing measures for the reference model and the currently measured - adequately processed - signal. The method was verified in the active experiment performed on the Fiat Punto automobile with a 5-speed gearbox. The described method can constitute the supplementation of on-board automobile system by the mechanical failures of the drive system. This procedure can be broadened by defects of the main gear, differential gear, articulated joints, bearing, engine exhaust valves, and engine head gasket.
Access status: Open Access ,
Semi-active suspension system modelling and parameters identification
(Wydawnictwa AGH, 2012) Kogut, Krzysztof
This paper presents a study of obtaining a model of the real quarter-car suspension device. The system is equipped with an automotive engineering magnetorheological (MR) rotary brake. Due to the complex mechanical structure of the apparatus the considered model contains several simplifications. In the parameter estimation process the grey-box method was used, while the process itself was split into two separate steps. In the first step, the parameters of the nonlinear model of suspension part are identified and the static profile of the MR damper is experimentally determined. The second step is to estimate the parameters of the model of the wheel-eccentricity part. Comparison of the modelled system trajectories and real-time experiments are presented. The identification results show that the obtained model is accurate and can be successfully applied to simulate the device.