Experimental and model-based analysis of motorcycle wobble and weave

This report presents analysis of motorcycle wobble and weave in different ways. The first method is the identification of wobble and weave via Stochastic Subspace Identification (SSI) and the second makes use of the software FastBike. Three SSI variants have been outlined and implemented in Matlab namely, Covariance (COV), Unweighted Principal Component (UPC) and the Conanical Variate Analysis (CVA). Next, pole selection using stabilisation diagrams is discussed and the different variants are compared. It is concluded that the covariance driven variant shows the most promising results. Finally, SSI-COV is used to identify weave and wobble of experimental data for different motorcycles. The second method presented in this report uses two models for comparing wobble and weave. The model by R.S. Sharp has been implemented in Fast- Bike and is compared to the original model. It has been concluded that the motorcycle stability model by Sharp is a good baseline because it corresponds to the results provided by the model implemented in FastBike.