Research On Response Of The Vehicle Under Multi-correlative Subsystems Excitations

With the development of automotive industry, vibration and noise problems of vehicles come out gradually, and have become a restriction for comfort. When vibration fault appears in a car, it may be caused by multi excitation source, and correlations exit in excitations of different grades. Nowadays, in vibration signals analysis of multi inputs, the research methods are usually using in analysizing the influence of some excitation source to final vibrated result, and it is incompleted. As a result, on the basis of feature extraction of subsystem inputs, it has important theoretical meaning and engineered significance.To address the vibration and noise problems of vehicles, the responses under multi subsystems extations have been researched and analyzed, which are summarized in the followings:(1) The research of traditional vibration and noise analysis methods have been reviewed, the main methods using in signal feature extraction of vehicle vibration and noise have been analyzed in detail, of which the disadvantages have also been pointed out.(2) A multi subsystem-inputs/single-output system model is proposed. On the basis of one grade input/single-output system model, the multi subsystem-inputs/single-output system model is introduced. After analysizing the relationship of correlated excitations, the practical system is simplified into a multi subsystem-inputs/single-output system model.(3) A new method using in feature extraction of multi subsystem-inputs/single-output system model is established. The Empirical Mode Decomposition method (EMD) has been improved after analysis. And combining the multi subsystem-inputs model, a new feature extraction method is established which based on multi subsystem-inputs model and Empirical Mode Decomposition method, and it provides conditions for the further contribution analysis of each input to response in subsystem-inputs and each subsystem-input to final response. At last, the simulation analysis verified its effectiveness.(4) A new method using in analysizing multi subsystem-inputs model is established. To address the shortage of partial coherence method exists in analysizing multi subsystem-inputs/single-output system model, the collective average method is used to analysize the system model, and after that, the multi correlated subsystem average inputs/single-output system model method is put forward, and the technique using in analysizing the contribution of correlated subsystem excitations to the whole system response is gained. The simulation analysis verified its effectiveness. (5) A whole vehicle test reseach is finished; the test is about the respongse under multi subsystem excitations. After the test of some vehicle is carried out on the real road, the new method using in feature extraction and new method using in analysizing multi subsystem-inputs model are used to analysized the signals collected in the test, and the reault provides evidence for the further improvement of the vehicle. The test result shows that, the new methods rosed in this paper can be applied in practical engineering problems efectively, and have certain valuable reference to the same project problems.In summary, to address the complex correlations exist in excitations of vehicle structure, theoretical analysis of multi correlated subsystem excitations is carried out, and test research is accomplished at last, a number of valuable conclusions have been got.