Dynamic Parameter Identification of a Bicycle Using Sliding Mode Observer and Particle Swarm Optimization

The aim of the present work is to identify the unknown dynamic parameters of a bicycle using a sliding mode observer and Particle Swarm Optimization (PSO) approaches. The estimation of bicycle dynamics requires a good knowledge of dynamic parameters such as damping coefficient, spring stiffness, and unsprung masses, among others. In this paper, suspension stiffness and damping coefficient have been identified using the Least Squares Method and compared with those obtained using the PSO technique. Real-time tests have been carried out on an instrumented bicycle equipped with various sensors to measure its dynamics. The measures coming from these sensors have been considered to validate the estimation tools. However, only the vertical wheel displacement measurements have been used to perform the observer. Experimental results are presented and discussed to demonstrate the quality of the proposed approach.