An Efficient PSO-Based Method for an Identification of a Quadrotor Model Parameters
[ 1 ] Instytut Automatyki i Inżynierii Informatycznej, Wydział Elektryczny, Politechnika Poznańska | [ D ] doktorant | [ P ] pracownik
2015
rozdział w monografii naukowej
angielski
- mathematical modeling
- identication
- particle swarm optimization
EN This paper considers the method of Quadrotor’s model parameters identification. Nowadays, restrictions are being imposed on the drons, forcing their control algorithms to be robust and faultless. This can be partially ensured by Model Reference Adaptive Control (MRAC) as well as dedicated state estimators (e.g. Extended Kalman Filter). Although those methods can be easy implemented and used, in all scenario, the parameterized model is needed. In this work we proposed the identification method for parameters of the quadrotor’s orientation model, based on the PSO (Particle Swarm Optimization). We also add different physical aspects to model, so it can characterize the real Quadrotor more precisely. The conducted experiments shows that the PSO, can provide fast and reliable estimation of the model parameters. It also reveals interesting nature of the proposed models.
95 - 104